One of the following — either one fixes the bug, pick whichever matches the project's case convention: 1. Make model lookup case-insensitive too. `--model anthropic/CLAUDE-OPUS-4-7` should resolve to the same model as `--model anthropic/claude-opus-4-7` (the catalog already does this for the provider half) and dispatch the canonical (lowercase) id to the provider. Result: both succeed. 2. Make model lookup case-sensitive AND normalize the catalog rejection. `--model anthropic/CLAUDE-OPUS-4-7` should fail with `Error: Unknown model: anthropic/CLAUDE-OPUS-4-7` (mirroring how `anthropic/claude-fake-9000` fails today), with no provider call. Optionally include a "did you mean: anthropic/claude-opus-4-7?" hint when a case-only difference is detected. Concrete grounded reference from the same CLI/build: - `--model anthropic/claude-fake-9000` (model that doesn't exist in any case) is rejected upfront with `Error: Unknown model: ...`, exit 1, no provider call. The Unknown-model error path exists. - `--model Anthropic/claude-opus-4-7`, `--model ANTHROPIC/claude-opus-4-7` (case-mismatched provider) both work — provider lookup is case-insensitive. The case-insensitive normalization exists. The bug is the gap between these two paths: the CLI normalizes provider case but not model case, and skips the unknown-model rejection for case-mismatched model ids.

openclaw - ✅(Solved) Fix [Bug]: --model anthropic/CLAUDE-OPUS-4-7 (case-mismatched model name) is accepted by CLI catalog and dispatched to the provider, surfacing as misleading "No text output returned" — provider name is case-insensitive but model name is not [1 pull requests, 2 comments, 3 participants]

iot2edge · 2026-04-28T17:55:18Z

[openclaw] pnpm openclaw infer model run --model / ` accepts the **provider** part case-insensitively (`anthropic`, `Anthropic`, `ANTHROPIC` all resolve to the same provider) but treats the **model** part case-sensitively at the provider call. The CLI catalog lookup does not reject case-mismatched model names; the request goes through to the provider with the mismatched id, which fails (or returns no content). The user sees `Error: No text output returned for provider "anthropic" model "CLAUDE-OPUS-4-7".` — the same misleading "No text output returned" wording from #73185 — when the actual cause is "model id case mismatch." Either model lookup should be case-insensitive (matching provider behavior), or the CLI should reject case-mismatched model ids upfront with a clear "Unknown model" error. # PR #73717: fix(cli): canonicalize infer model run refs - Repository: openclaw/openclaw - Author: ai-hpc - State: open | merged: False - Link: https://github.com/openclaw/openclaw/pull/73717 ## Description (problem / solution / changelog) ## Summary - Problem: `infer model run --model` forwarded case-mismatched catalog model ids such as `Anthropic/CLAUDE-OPUS-4-7` to the model runtime, which could surface as a misleading empty-output provider error. - Why it matters: users can paste natural mixed-case model refs and get a provider-looking failure instead of the canonical known model behavior. - What changed: explicit model-run overrides now resolve through the model catalog when there is a unique case-insensitive provider/model match, then dispatch the canonical provider/model ref for both local and gateway transports. - What did NOT change (scope boundary): custom mixed-case refs are preserved when no unique catalog match exists; this does not change model catalog listing, provider auth, or non-`infer model run` commands. ## Change Type (select all) - [x] Bug fix - [ ] Feature - [ ] Refactor required for the fix - [ ] Docs - [ ] Security hardening - [ ] Chore/infra ## Scope (select all touched areas) - [ ] Gateway / orchestration - [ ] Skills / tool execution - [ ] Auth / tokens - [ ] Memory / storage - [ ] Integrations - [ ] API / contracts - [x] UI / DX - [ ] CI/CD / infra ## Linked Issue/PR - Closes #73715 - [x] This PR fixes a bug or regression ## Root Cause (if applicable) - Root cause: `infer model run` parsed and forwarded explicit model overrides without canonicalizing case-only matches against the known model catalog. - Missing detection / guardrail: no regression test covered case-mismatched explicit model refs before local or gateway dispatch. - Contributing context (if known): provider ids are normalized case-insensitively, while model ids were forwarded with user-supplied casing. ## Regression Test Plan (if applicable) - Coverage level that should have caught this: - [x] Unit test - [x] Seam / integration test - [ ] End-to-end test - [ ] Existing coverage already sufficient - Target test or file: `src/cli/capability-cli.test.ts` - Scenario the test should lock in: `Anthropic/CLAUDE-OPUS-4-7` is canonicalized to `anthropic/claude-opus-4-7` before local and gateway model-run dispatch. - Why this is the smallest reliable guardrail: the bug is in the CLI override-to-dispatch seam, so mocked dispatch tests catch the raw value before any provider call. - Existing test that already covers this (if any): none. - If no new test is added, why not: N/A. ## User-visible / Behavior Changes `openclaw infer model run --model ` now accepts case-only mismatches for catalog models when the match is unique, dispatching the canonical model id instead of the raw casing. ## Diagram (if applicable) ```text Before: --model Anthropic/CLAUDE-OPUS-4-7 -> raw runtime dispatch -> misleading provider/no-output error After: --model Anthropic/CLAUDE-OPUS-4-7 -> unique catalog match -> anthropic/claude-opus-4-7 -> normal dispatch ``` ## Security Impact (required) - New permissions/capabilities? (`Yes/No`): No - Secrets/tokens handling changed? (`Yes/No`): No - New/changed network calls? (`Yes/No`): No - Command/tool execution surface changed? (`Yes/No`): No - Data access scope changed? (`Yes/No`): No - If any `Yes`, explain risk + mitigation: N/A ## Repro + Verification ### Environment - OS: Ubuntu 24.04.4 LTS - Runtime/container: Node 22, pnpm dev checkout - Model/provider: Anthropic catalog model refs - Integration/channel (if any): CLI `infer model run` - Relevant config (redacted): default source-checkout runtime config; no secrets included ### Steps 1. Run `openclaw infer model run --model Anthropic/CLAUDE-OPUS-4-7 --prompt "hello" --json`. 2. Observe the model-run dispatch receives the canonical ref. 3. Run the same path with a custom mixed-case ref that has no catalog match. ### Expected - Known catalog model refs with case-only mismatches dispatch a

openclaw2026-04-28 17:55:18

ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

GitHub issue URL

Helpful · Quick feedback

GitHub stats

openclaw/openclaw#73715•Fetched 2026-04-29 06:16:03

View on GitHub

Comments

Participants

Timeline

Reactions

Author

Participants

Timeline (top)

commented ×2cross-referenced ×2labeled ×1mentioned ×1

pnpm openclaw infer model run --model <provider>/<model> accepts the provider part case-insensitively (anthropic, Anthropic, ANTHROPIC all resolve to the same provider) but treats the model part case-sensitively at the provider call. The CLI catalog lookup does not reject case-mismatched model names; the request goes through to the provider with the mismatched id, which fails (or returns no content). The user sees Error: No text output returned for provider "anthropic" model "CLAUDE-OPUS-4-7". — the same misleading "No text output returned" wording from #73185 — when the actual cause is "model id case mismatch." Either model lookup should be case-insensitive (matching provider behavior), or the CLI should reject case-mismatched model ids upfront with a clear "Unknown model" error.

Error Message

$ pnpm openclaw infer model run --model anthropic/CLAUDE-OPUS-4-7 --prompt "Just say OK" Error: No text output returned for provider "anthropic" model "CLAUDE-OPUS-4-7". ELIFECYCLE Command failed with exit code 1.

$ pnpm openclaw infer model run --model anthropic/Claude-Opus-4-7 --prompt "Just say OK" Error: No text output returned for provider "anthropic" model "Claude-Opus-4-7". ELIFECYCLE Command failed with exit code 1.

$ pnpm openclaw infer model run --model anthropic/CLAUDE-OPUS-4-7 --prompt "Just say OK" --json Error: No text output returned for provider "anthropic" model "CLAUDE-OPUS-4-7". ELIFECYCLE Command failed with exit code 1.

Root Cause

The error message reports the exact case the user typed — proving the model id was passed through to the provider unchanged. Note that the "No text output returned" wording is the same one #73185 patches — that issue is about empty --prompt; this one is about case-mismatched --model. Both have the same downstream symptom for the same root cause: input not normalized/validated before dispatch.

Fix Action

Fix / Workaround

Make model lookup case-insensitive too. --model anthropic/CLAUDE-OPUS-4-7 should resolve to the same model as --model anthropic/claude-opus-4-7 (the catalog already does this for the provider half) and dispatch the canonical (lowercase) id to the provider. Result: both succeed.
Make model lookup case-sensitive AND normalize the catalog rejection. --model anthropic/CLAUDE-OPUS-4-7 should fail with Error: Unknown model: anthropic/CLAUDE-OPUS-4-7 (mirroring how anthropic/claude-fake-9000 fails today), with no provider call. Optionally include a "did you mean: anthropic/claude-opus-4-7?" hint when a case-only difference is detected.

Likely fix locus: the model-id normalization/lookup path (search for "Unknown model:" in src/cli/cli-infer/ or wherever the parser lives in this build, plus the model-catalog resolution helper). Two viable approaches:
1. Match provider behavior — make model lookup case-insensitive, normalize the matched canonical id, and dispatch the canonical id to the provider. Most user-friendly.
2. Make model lookup case-sensitive AND route case-mismatched ids through the existing "Unknown model" rejection path. Most consistent with literal id semantics. Optionally add a "did you mean: ...?" hint.

PR fix notes

PR #73717: fix(cli): canonicalize infer model run refs

Repository: openclaw/openclaw
Author: ai-hpc
State: open | merged: False
Link: https://github.com/openclaw/openclaw/pull/73717

Description (problem / solution / changelog)

Summary

Problem: infer model run --model forwarded case-mismatched catalog model ids such as Anthropic/CLAUDE-OPUS-4-7 to the model runtime, which could surface as a misleading empty-output provider error.
Why it matters: users can paste natural mixed-case model refs and get a provider-looking failure instead of the canonical known model behavior.
What changed: explicit model-run overrides now resolve through the model catalog when there is a unique case-insensitive provider/model match, then dispatch the canonical provider/model ref for both local and gateway transports.
What did NOT change (scope boundary): custom mixed-case refs are preserved when no unique catalog match exists; this does not change model catalog listing, provider auth, or non-infer model run commands.

Change Type (select all)

Scope (select all touched areas)

Linked Issue/PR

Closes #73715
This PR fixes a bug or regression

Root Cause (if applicable)

Root cause: infer model run parsed and forwarded explicit model overrides without canonicalizing case-only matches against the known model catalog.
Missing detection / guardrail: no regression test covered case-mismatched explicit model refs before local or gateway dispatch.
Contributing context (if known): provider ids are normalized case-insensitively, while model ids were forwarded with user-supplied casing.

Regression Test Plan (if applicable)

Coverage level that should have caught this:
- Unit test
- Seam / integration test
- End-to-end test
- Existing coverage already sufficient
Target test or file: src/cli/capability-cli.test.ts
Scenario the test should lock in: Anthropic/CLAUDE-OPUS-4-7 is canonicalized to anthropic/claude-opus-4-7 before local and gateway model-run dispatch.
Why this is the smallest reliable guardrail: the bug is in the CLI override-to-dispatch seam, so mocked dispatch tests catch the raw value before any provider call.
Existing test that already covers this (if any): none.
If no new test is added, why not: N/A.

User-visible / Behavior Changes

openclaw infer model run --model <provider/model> now accepts case-only mismatches for catalog models when the match is unique, dispatching the canonical model id instead of the raw casing.

Diagram (if applicable)

Before:
--model Anthropic/CLAUDE-OPUS-4-7 -> raw runtime dispatch -> misleading provider/no-output error

After:
--model Anthropic/CLAUDE-OPUS-4-7 -> unique catalog match -> anthropic/claude-opus-4-7 -> normal dispatch

Security Impact (required)

New permissions/capabilities? (Yes/No): No
Secrets/tokens handling changed? (Yes/No): No
New/changed network calls? (Yes/No): No
Command/tool execution surface changed? (Yes/No): No
Data access scope changed? (Yes/No): No
If any Yes, explain risk + mitigation: N/A

Repro + Verification

Environment

OS: Ubuntu 24.04.4 LTS
Runtime/container: Node 22, pnpm dev checkout
Model/provider: Anthropic catalog model refs
Integration/channel (if any): CLI infer model run
Relevant config (redacted): default source-checkout runtime config; no secrets included

Steps

Run openclaw infer model run --model Anthropic/CLAUDE-OPUS-4-7 --prompt "hello" --json.
Observe the model-run dispatch receives the canonical ref.
Run the same path with a custom mixed-case ref that has no catalog match.

Expected

Known catalog model refs with case-only mismatches dispatch as their canonical provider/model id.
Custom mixed-case refs without a unique catalog match remain unchanged.

Actual

Before this PR, the raw mixed-case catalog model id was forwarded.
After this PR, the CLI canonicalizes only unique catalog matches before dispatch.

Evidence

Failing test/log before + passing after
Trace/log snippets
Screenshot/recording
Perf numbers (if relevant)

pnpm test src/cli/capability-cli.test.ts -- --reporter=verbose
Test Files  1 passed (1)
Tests  47 passed (47)

pnpm check:changed
passed

git diff --check
passed

Human Verification (required)

Verified scenarios: local dispatch canonicalization, gateway dispatch canonicalization, custom mixed-case ref preservation.
Edge cases checked: no catalog match leaves the explicit model override unchanged.
What you did not verify: live Anthropic provider call.
pnpm test src/cli/capability-cli.test.ts -- --reporter=verbose passed: 51 tests.
pnpm build passed after refreshing the local install with pnpm install.
Live CLI smoke with existing local OpenRouter auth profile passed:
- Command shape: pnpm --silent openclaw infer model run --model OpenRouter/OPENROUTER/AUTO --prompt 'Reply with OK only.' --json
- Result: exit 0, parseable JSON, ok=true, provider=openrouter, model=openrouter/auto, output text OK.
Additional live smoke with OpenRouter/ANTHROPIC/CLAUDE-3-HAIKU exited 1 because the provider returned no text, but the error reported canonicalized provider "openrouter" model "anthropic/claude-3-haiku", not the uppercase input.

Review Conversations

I replied to or resolved every bot review conversation I addressed in this PR.
I left unresolved only the conversations that still need reviewer or maintainer judgment.

Compatibility / Migration

Backward compatible? (Yes/No): Yes
Config/env changes? (Yes/No): No
Migration needed? (Yes/No): No
If yes, exact upgrade steps: N/A

Risks and Mitigations

Risk: A custom provider intentionally using mixed-case ids could be changed unexpectedly.
- Mitigation: canonicalization only happens when the loaded model catalog has exactly one case-insensitive provider/model match; otherwise the raw ref is preserved.

Changed files

src/cli/capability-cli.test.ts (modified, +110/-0)
src/cli/capability-cli.ts (modified, +64/-5)

Code Example

$ pnpm openclaw infer model run --model anthropic/CLAUDE-OPUS-4-7 --prompt "Just say OK"
Error: No text output returned for provider "anthropic" model "CLAUDE-OPUS-4-7".
 ELIFECYCLE  Command failed with exit code 1.

$ pnpm openclaw infer model run --model anthropic/Claude-Opus-4-7 --prompt "Just say OK"
Error: No text output returned for provider "anthropic" model "Claude-Opus-4-7".
 ELIFECYCLE  Command failed with exit code 1.

$ pnpm openclaw infer model run --model anthropic/CLAUDE-OPUS-4-7 --prompt "Just say OK" --json
Error: No text output returned for provider "anthropic" model "CLAUDE-OPUS-4-7".
 ELIFECYCLE  Command failed with exit code 1.

---

- Default model: anthropic/claude-opus-4-7 (set via `pnpm openclaw models set anthropic/claude-opus-4-7`).
- Anthropic auth: Claude CLI OAuth (detected by `openclaw doctor`).
- No per-agent overrides; default agent `main`; gateway `mode=local`.
- The bug reproduces under both `--local` (default in this run) and `--gateway` transports — this is upstream of transport selection, in the model-id normalization path.

---

=== Verified case-sensitivity matrix (single shell session, OpenClaw 2026.4.27 / 0450bba, Node 22.22.2) ===

EXIT  MODEL ID                                       RESULT
0     anthropic/claude-opus-4-7                      outputs: 1                                      (canonical case)
0     Anthropic/claude-opus-4-7                      outputs: 1                                      (provider uppercase OK)
0     ANTHROPIC/claude-opus-4-7                      outputs: 1                                      (provider full uppercase OK)
1     anthropic/Claude-Opus-4-7                      Error: No text output returned for provider "anthropic" model "Claude-Opus-4-7"
1     anthropic/CLAUDE-OPUS-4-7                      Error: No text output returned for provider "anthropic" model "CLAUDE-OPUS-4-7"
1     Anthropic/CLAUDE-OPUS-4-7                      Error: No text output returned for provider "anthropic" model "CLAUDE-OPUS-4-7"

=== The bug: case-mismatched model name reaches the provider ===

$ pnpm openclaw infer model run --model anthropic/CLAUDE-OPUS-4-7 --prompt "Just say OK"
Error: No text output returned for provider "anthropic" model "CLAUDE-OPUS-4-7".
 ELIFECYCLE  Command failed with exit code 1.

$ pnpm openclaw infer model run --model anthropic/CLAUDE-OPUS-4-7 --prompt "Just say OK" --json
Error: No text output returned for provider "anthropic" model "CLAUDE-OPUS-4-7".
 ELIFECYCLE  Command failed with exit code 1.

=== In-tree counter-example #1: provider name is case-insensitive ===

$ pnpm openclaw infer model run --model Anthropic/claude-opus-4-7 --prompt "Just say OK" --json | jq -r '.outputs[0].text'
OK

$ pnpm openclaw infer model run --model ANTHROPIC/claude-opus-4-7 --prompt "Just say OK" --json | jq -r '.outputs[0].text'
OK

=== In-tree counter-example #2: nonexistent model is rejected upfront ===

$ pnpm openclaw infer model run --model anthropic/claude-fake-9000 --prompt "Just say OK"
Error: Unknown model: anthropic/claude-fake-9000
 ELIFECYCLE  Command failed with exit code 1.

(Note: this rejection happens BEFORE any provider call. The Unknown-model code path exists; case-mismatched model ids should take the same path.)

=== Sanity check: the canonical lowercase model id works correctly ===

$ pnpm openclaw infer model run --model anthropic/claude-opus-4-7 --prompt "Just say OK" --json | jq -r '.outputs[0].text'
OK

---

Affected users/systems/channels:
- Every operator who pastes a model id with mixed case (e.g. copy-paste from docs or chat where headers are title-cased) into `infer model run --model`. Linux directly observed (Ubuntu 24.04 / Node 22.22.2 / pnpm 10.33.0); platform-agnostic code path so macOS/Windows are expected to reproduce.
- New users who don't know the canonical lowercase form of model ids and try natural-feeling capitalizations (`Claude-Opus-4-7`, `CLAUDE-OPUS-4-7`).
- Users hitting this on Anthropic specifically; whether other providers exhibit the same asymmetry depends on each provider plugin's normalization (not directly verified for non-anthropic providers).

Severity:
- Annoying with confusing diagnostics. The "No text output returned" error misattributes a CLI input mistake to a provider/model fault, sending users down the wrong debugging path.
- For non-Claude providers that bill on partial requests: the case-mismatched call may incur tokens depending on whether the provider rejects pre-call or accepts and returns empty (not directly verified for non-anthropic providers).
- Inconsistent with the in-tree pattern. Provider name is case-insensitive; model name is not. Same input field, two case-sensitivity rules.
- Not a security/data-loss bug, no crash.

Frequency:
- Always, deterministic for case-mismatched anthropic model names. 100% reproduction across `Claude-Opus-4-7`, `CLAUDE-OPUS-4-7`, and `Anthropic/CLAUDE-OPUS-4-7`. Independent of model/transport.

Consequence:
- Operators waste time debugging "why doesn't claude work?" when the real issue is case mismatch in the model id.
- Inconsistency erodes operator trust in input parsing — the CLI accepts case-insensitive provider but rejects (downstream, with a misleading error) case-mismatched model.
- The "No text output returned" error string is now overloaded across at least three failure modes: empty prompt (#73185), case-mismatched model (this), and a configured provider returning empty body for valid input (closed #65394, #66506, #65076). Every overload makes the error less useful.
- No grounded evidence of missed messages, failed onboarding, or extra cost.

---

- Regression status: not classified as a Regression. Last-known-good not directly observed; no bisect performed.

- Likely fix locus: the model-id normalization/lookup path (search for "Unknown model:" in `src/cli/cli-infer/` or wherever the parser lives in this build, plus the model-catalog resolution helper). Two viable approaches:
  1. Match provider behavior — make model lookup case-insensitive, normalize the matched canonical id, and dispatch the canonical id to the provider. Most user-friendly.
  2. Make model lookup case-sensitive AND route case-mismatched ids through the existing "Unknown model" rejection path. Most consistent with literal id semantics. Optionally add a "did you mean: ...?" hint.

- Suggested regression test: a unit test on the model-id resolver that asserts each of `Anthropic/claude-opus-4-7`, `anthropic/CLAUDE-OPUS-4-7`, `anthropic/Claude-Opus-4-7`, `Anthropic/CLAUDE-OPUS-4-7` either (a) resolves to the canonical `anthropic/claude-opus-4-7` and dispatches accordingly, OR (b) is rejected with `Error: Unknown model: <input>`. Pair with an existing-behavior parity test that the canonical lowercase id still works.

- Related findings:
  - #73185 (filed): empty `--prompt` reaches the provider on `--local` transport. Same misleading downstream error string ("No text output returned"), different upstream cause (empty content vs case-mismatched model). Same root-cause class: the CLI doesn't validate/normalize input before dispatching.
  - Closed issues #65394, #66506, #65076: addressed "No text output returned" in valid-input scenarios. The same string remains overloaded across at least three failure modes; this issue makes that overload one mode larger.
  - Related parsing edge case: `--model anthropic/` (provider with trailing slash, no model) currently resolves to the provider's default model and succeeds — undocumented hidden behavior. Could be filed separately if maintainers consider it a UX gap.
  - Related parsing edge case: `--model anthropic` (provider only, no slash) currently fails with `Error: Unknown model: anthropic/anthropic` — implicit `<provider>/<provider>` auto-prefix exposes internal logic in the error message. Could be filed separately.

- Dedupe checked against the openclaw issue corpus on 2026-04-28: no existing open or closed issue matches the case-sensitivity asymmetry pattern specifically. Searched "case sensitive model id", "model id case mismatch", "case insensitive provider model", "uppercase model name infer" — zero direct matches.

- Not exercised in this repro: case-sensitivity behavior for non-anthropic providers (deepseek, openrouter, etc.); behavior under `--gateway` transport (the bug is in the model-id resolver upstream of transport, but not directly verified on gateway path); aliases with case mismatches (e.g. `Opus` vs `opus`).

RAW_BUFFERClick to expand / collapse

Bug type

Behavior bug (incorrect output/state without crash)

Beta release blocker

Summary

Steps to reproduce

Fresh checkout of openclaw at v2026.4.27 (commit 0450bba); pnpm install && pnpm build on Node 22.22.2.
With anthropic auth configured (Claude CLI OAuth in this run).
Run pnpm openclaw infer model run --model anthropic/CLAUDE-OPUS-4-7 --prompt "Just say OK".
Observe Error: No text output returned for provider "anthropic" model "CLAUDE-OPUS-4-7". and exit 1. The catalog accepted the case-mismatched id; the provider call ran and failed; the error misattributes the cause.
Run with mixed case in the model: --model anthropic/Claude-Opus-4-7. Same misleading error.
Counter-example for the provider half of the id: --model Anthropic/claude-opus-4-7, --model ANTHROPIC/claude-opus-4-7. Both succeed with exit 0 — provider name is case-insensitive.
Counter-example for an actually-unknown model: --model anthropic/claude-fake-9000. CLI rejects upfront with Error: Unknown model: anthropic/claude-fake-9000 and exit 1, before any provider call. The catalog has the correct error path; it just doesn't take the case-mismatched route.

Expected behavior

One of the following — either one fixes the bug, pick whichever matches the project's case convention:

Make model lookup case-insensitive too. --model anthropic/CLAUDE-OPUS-4-7 should resolve to the same model as --model anthropic/claude-opus-4-7 (the catalog already does this for the provider half) and dispatch the canonical (lowercase) id to the provider. Result: both succeed.
Make model lookup case-sensitive AND normalize the catalog rejection. --model anthropic/CLAUDE-OPUS-4-7 should fail with Error: Unknown model: anthropic/CLAUDE-OPUS-4-7 (mirroring how anthropic/claude-fake-9000 fails today), with no provider call. Optionally include a "did you mean: anthropic/claude-opus-4-7?" hint when a case-only difference is detected.

Concrete grounded reference from the same CLI/build:

--model anthropic/claude-fake-9000 (model that doesn't exist in any case) is rejected upfront with Error: Unknown model: ..., exit 1, no provider call. The Unknown-model error path exists.
--model Anthropic/claude-opus-4-7, --model ANTHROPIC/claude-opus-4-7 (case-mismatched provider) both work — provider lookup is case-insensitive. The case-insensitive normalization exists.

The bug is the gap between these two paths: the CLI normalizes provider case but not model case, and skips the unknown-model rejection for case-mismatched model ids.

Actual behavior

$ pnpm openclaw infer model run --model anthropic/CLAUDE-OPUS-4-7 --prompt "Just say OK"
Error: No text output returned for provider "anthropic" model "CLAUDE-OPUS-4-7".
 ELIFECYCLE  Command failed with exit code 1.

$ pnpm openclaw infer model run --model anthropic/Claude-Opus-4-7 --prompt "Just say OK"
Error: No text output returned for provider "anthropic" model "Claude-Opus-4-7".
 ELIFECYCLE  Command failed with exit code 1.

$ pnpm openclaw infer model run --model anthropic/CLAUDE-OPUS-4-7 --prompt "Just say OK" --json
Error: No text output returned for provider "anthropic" model "CLAUDE-OPUS-4-7".
 ELIFECYCLE  Command failed with exit code 1.

OpenClaw version

2026.4.27

Operating system

Ubuntu 24.04.4

Install method

pnpm dev

Model

anthropic/claude-opus-4-7

Provider / routing chain

openclaw -> local transport -> anthropic

Additional provider/model setup details

- Default model: anthropic/claude-opus-4-7 (set via `pnpm openclaw models set anthropic/claude-opus-4-7`).
- Anthropic auth: Claude CLI OAuth (detected by `openclaw doctor`).
- No per-agent overrides; default agent `main`; gateway `mode=local`.
- The bug reproduces under both `--local` (default in this run) and `--gateway` transports — this is upstream of transport selection, in the model-id normalization path.

Logs, screenshots, and evidence

=== Verified case-sensitivity matrix (single shell session, OpenClaw 2026.4.27 / 0450bba, Node 22.22.2) ===

EXIT  MODEL ID                                       RESULT
0     anthropic/claude-opus-4-7                      outputs: 1                                      (canonical case)
0     Anthropic/claude-opus-4-7                      outputs: 1                                      (provider uppercase OK)
0     ANTHROPIC/claude-opus-4-7                      outputs: 1                                      (provider full uppercase OK)
1     anthropic/Claude-Opus-4-7                      Error: No text output returned for provider "anthropic" model "Claude-Opus-4-7"
1     anthropic/CLAUDE-OPUS-4-7                      Error: No text output returned for provider "anthropic" model "CLAUDE-OPUS-4-7"
1     Anthropic/CLAUDE-OPUS-4-7                      Error: No text output returned for provider "anthropic" model "CLAUDE-OPUS-4-7"

=== The bug: case-mismatched model name reaches the provider ===

$ pnpm openclaw infer model run --model anthropic/CLAUDE-OPUS-4-7 --prompt "Just say OK"
Error: No text output returned for provider "anthropic" model "CLAUDE-OPUS-4-7".
 ELIFECYCLE  Command failed with exit code 1.

$ pnpm openclaw infer model run --model anthropic/CLAUDE-OPUS-4-7 --prompt "Just say OK" --json
Error: No text output returned for provider "anthropic" model "CLAUDE-OPUS-4-7".
 ELIFECYCLE  Command failed with exit code 1.

=== In-tree counter-example #1: provider name is case-insensitive ===

$ pnpm openclaw infer model run --model Anthropic/claude-opus-4-7 --prompt "Just say OK" --json | jq -r '.outputs[0].text'
OK

$ pnpm openclaw infer model run --model ANTHROPIC/claude-opus-4-7 --prompt "Just say OK" --json | jq -r '.outputs[0].text'
OK

=== In-tree counter-example #2: nonexistent model is rejected upfront ===

$ pnpm openclaw infer model run --model anthropic/claude-fake-9000 --prompt "Just say OK"
Error: Unknown model: anthropic/claude-fake-9000
 ELIFECYCLE  Command failed with exit code 1.

(Note: this rejection happens BEFORE any provider call. The Unknown-model code path exists; case-mismatched model ids should take the same path.)

=== Sanity check: the canonical lowercase model id works correctly ===

$ pnpm openclaw infer model run --model anthropic/claude-opus-4-7 --prompt "Just say OK" --json | jq -r '.outputs[0].text'
OK

Impact and severity

Affected users/systems/channels:
- Every operator who pastes a model id with mixed case (e.g. copy-paste from docs or chat where headers are title-cased) into `infer model run --model`. Linux directly observed (Ubuntu 24.04 / Node 22.22.2 / pnpm 10.33.0); platform-agnostic code path so macOS/Windows are expected to reproduce.
- New users who don't know the canonical lowercase form of model ids and try natural-feeling capitalizations (`Claude-Opus-4-7`, `CLAUDE-OPUS-4-7`).
- Users hitting this on Anthropic specifically; whether other providers exhibit the same asymmetry depends on each provider plugin's normalization (not directly verified for non-anthropic providers).

Severity:
- Annoying with confusing diagnostics. The "No text output returned" error misattributes a CLI input mistake to a provider/model fault, sending users down the wrong debugging path.
- For non-Claude providers that bill on partial requests: the case-mismatched call may incur tokens depending on whether the provider rejects pre-call or accepts and returns empty (not directly verified for non-anthropic providers).
- Inconsistent with the in-tree pattern. Provider name is case-insensitive; model name is not. Same input field, two case-sensitivity rules.
- Not a security/data-loss bug, no crash.

Frequency:
- Always, deterministic for case-mismatched anthropic model names. 100% reproduction across `Claude-Opus-4-7`, `CLAUDE-OPUS-4-7`, and `Anthropic/CLAUDE-OPUS-4-7`. Independent of model/transport.

Consequence:
- Operators waste time debugging "why doesn't claude work?" when the real issue is case mismatch in the model id.
- Inconsistency erodes operator trust in input parsing — the CLI accepts case-insensitive provider but rejects (downstream, with a misleading error) case-mismatched model.
- The "No text output returned" error string is now overloaded across at least three failure modes: empty prompt (#73185), case-mismatched model (this), and a configured provider returning empty body for valid input (closed #65394, #66506, #65076). Every overload makes the error less useful.
- No grounded evidence of missed messages, failed onboarding, or extra cost.

Additional information

- Regression status: not classified as a Regression. Last-known-good not directly observed; no bisect performed.

- Likely fix locus: the model-id normalization/lookup path (search for "Unknown model:" in `src/cli/cli-infer/` or wherever the parser lives in this build, plus the model-catalog resolution helper). Two viable approaches:
  1. Match provider behavior — make model lookup case-insensitive, normalize the matched canonical id, and dispatch the canonical id to the provider. Most user-friendly.
  2. Make model lookup case-sensitive AND route case-mismatched ids through the existing "Unknown model" rejection path. Most consistent with literal id semantics. Optionally add a "did you mean: ...?" hint.

- Suggested regression test: a unit test on the model-id resolver that asserts each of `Anthropic/claude-opus-4-7`, `anthropic/CLAUDE-OPUS-4-7`, `anthropic/Claude-Opus-4-7`, `Anthropic/CLAUDE-OPUS-4-7` either (a) resolves to the canonical `anthropic/claude-opus-4-7` and dispatches accordingly, OR (b) is rejected with `Error: Unknown model: <input>`. Pair with an existing-behavior parity test that the canonical lowercase id still works.

- Related findings:
  - #73185 (filed): empty `--prompt` reaches the provider on `--local` transport. Same misleading downstream error string ("No text output returned"), different upstream cause (empty content vs case-mismatched model). Same root-cause class: the CLI doesn't validate/normalize input before dispatching.
  - Closed issues #65394, #66506, #65076: addressed "No text output returned" in valid-input scenarios. The same string remains overloaded across at least three failure modes; this issue makes that overload one mode larger.
  - Related parsing edge case: `--model anthropic/` (provider with trailing slash, no model) currently resolves to the provider's default model and succeeds — undocumented hidden behavior. Could be filed separately if maintainers consider it a UX gap.
  - Related parsing edge case: `--model anthropic` (provider only, no slash) currently fails with `Error: Unknown model: anthropic/anthropic` — implicit `<provider>/<provider>` auto-prefix exposes internal logic in the error message. Could be filed separately.

- Dedupe checked against the openclaw issue corpus on 2026-04-28: no existing open or closed issue matches the case-sensitivity asymmetry pattern specifically. Searched "case sensitive model id", "model id case mismatch", "case insensitive provider model", "uppercase model name infer" — zero direct matches.

- Not exercised in this repro: case-sensitivity behavior for non-anthropic providers (deepseek, openrouter, etc.); behavior under `--gateway` transport (the bug is in the model-id resolver upstream of transport, but not directly verified on gateway path); aliases with case mismatches (e.g. `Opus` vs `opus`).

extent analysis

TL;DR

To fix the bug, make the model lookup case-insensitive by normalizing the input model id to lowercase before dispatching it to the provider.

Guidance

Identify the model-id normalization/lookup path in the codebase, likely in src/cli/cli-infer/, and modify it to normalize the input model id to lowercase.
Update the model catalog resolution helper to handle case-insensitive lookups.
Consider adding a "did you mean: ..." hint for case-mismatched model ids.
Write unit tests to verify the fix, including test cases for different casing scenarios.

Example

// Example of normalizing the model id to lowercase
const modelName = 'CLAUDE-OPUS-4-7';
const normalizedModelName = modelName.toLowerCase();
// Dispatch the normalized model name to the provider

Notes

The fix should be applied to the model-id resolver to ensure consistent behavior across different providers.
The existing "Unknown model" rejection path can be reused to handle case-mismatched model ids.

Recommendation

Apply the workaround by making the model lookup case-insensitive, as it is the most user-friendly approach and consistent with the provider's case-insensitive behavior.

Vote matrix · Quick signals

Works

Did the solution work? Tap to confirm.

Easy Fix

Was it a quick fix?

Time Saver

Did it save you time?

Blocking

Was it severely blocking?

Common Issue

Are others likely hitting this too?

Flaky / Intermittent

Is it intermittent?

Verified / Reproducible

Can you reproduce it reliably?

FAQ

Expected behavior

One of the following — either one fixes the bug, pick whichever matches the project's case convention:

Make model lookup case-insensitive too. --model anthropic/CLAUDE-OPUS-4-7 should resolve to the same model as --model anthropic/claude-opus-4-7 (the catalog already does this for the provider half) and dispatch the canonical (lowercase) id to the provider. Result: both succeed.
Make model lookup case-sensitive AND normalize the catalog rejection. --model anthropic/CLAUDE-OPUS-4-7 should fail with Error: Unknown model: anthropic/CLAUDE-OPUS-4-7 (mirroring how anthropic/claude-fake-9000 fails today), with no provider call. Optionally include a "did you mean: anthropic/claude-opus-4-7?" hint when a case-only difference is detected.

Concrete grounded reference from the same CLI/build:

--model anthropic/claude-fake-9000 (model that doesn't exist in any case) is rejected upfront with Error: Unknown model: ..., exit 1, no provider call. The Unknown-model error path exists.
--model Anthropic/claude-opus-4-7, --model ANTHROPIC/claude-opus-4-7 (case-mismatched provider) both work — provider lookup is case-insensitive. The case-insensitive normalization exists.

The bug is the gap between these two paths: the CLI normalizes provider case but not model case, and skips the unknown-model rejection for case-mismatched model ids.

#api #embedding generation #cache error #pipeline error #runtime error

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Recommended Tools

GitHub issue graph ai analysis

Error Message

Root Cause

Fix Action

Fix / Workaround

PR fix notes

PR #73717: fix(cli): canonicalize infer model run refs

Description (problem / solution / changelog)

Summary

Change Type (select all)

Scope (select all touched areas)

Linked Issue/PR

Root Cause (if applicable)

Regression Test Plan (if applicable)

User-visible / Behavior Changes

Diagram (if applicable)

Security Impact (required)

Repro + Verification

Environment

Steps

Expected

Actual

Evidence

Human Verification (required)

Review Conversations

Compatibility / Migration

Risks and Mitigations

Changed files

Code Example

Bug type

Beta release blocker

Summary

Steps to reproduce

Expected behavior

Actual behavior

OpenClaw version

Operating system

Install method

Model

Provider / routing chain

Additional provider/model setup details

Logs, screenshots, and evidence

Impact and severity

Additional information

extent analysis

TL;DR

Guidance

Example

Notes

Recommendation

FAQ

Expected behavior

Still need to ship something?

RELATED_DISCOVERY

TRENDING