Repository navigation
feat(opencode-go): register 4 missing models from upstream catalog - #2790
Merged
diegosouzapw merged 2 commits intoMay 27, 2026
Conversation
Add qwen3.7-max, mimo-v2-pro, mimo-v2-omni, hy3-preview to the opencode-go provider. Extend executor tests and model specs. - Registry now matches live https://opencode.ai/zen/go/v1/models response (16 models). Previously, four models returned by the upstream API were absent from the static registry, so requests to them failed model resolution before reaching the executor. - qwen3.7-max: inherits defaultContextLength 200000; model spec mirrors qwen3.6-plus (Bailian multimodal, thinking + tools + vision). - mimo-v2-pro: spec mirrors mimo-v2-omni (262144 ctx, 131072 maxOut, tools + vision). - mimo-v2-omni: registry entry only (spec already present upstream). - hy3-preview: registry entry only; falls back to __default__ spec. - All four route to /chat/completions (default openai format), matching the existing opencode-go executor behavior.
|
Kilo Code Review could not run — your account is out of credits. Add credits or switch to a free model to enable reviews on this change. |
Merged
2 of 5 tasks
diegosouzapw
merged commit May 27, 2026
3c016ce
into
cheaper-inference:release/v3.8.5
1 of 2 checks passed
jeferssonlemes
added a commit
to jeferssonlemes/OmniRoute
that referenced
this pull request
May 27, 2026
Upstream returns 401 "Model qwen3.x-* is not supported for format oa-compat" when hitting /chat/completions for qwen3.5-plus / qwen3.6-plus / qwen3.7-max on opencode-go. Same upstream behavior already documented for opencode-zen via Issue cheaper-inference#2292. - Adds `targetFormat: "claude"` to the three qwen entries in the opencode-go provider registry so requests route through /messages with the anthropic-version header. - Replaces the previous `qwen3.5-plus` / `qwen3.6-plus` rows in the PR cheaper-inference#1510 executor test (they registered the models without targetFormat, which now conflicts with the registry default). - Adds a new test asserting all three qwen ids route to /zen/go/v1/messages and carry the anthropic-version: 2023-06-01 header. Fork-only commit on chatlabel/patches — upstream PR cheaper-inference#2790 is still open and does not yet include this fix.
Merged
HouMinXi
pushed a commit
to HouMinXi/OmniRoute
that referenced
this pull request
Aug 2, 2026
…encode-go-missing-models feat(opencode-go): register 4 missing models from upstream catalog
Poid-ZA
pushed a commit
to Poid-ZA/OmniRoute
that referenced
this pull request
Aug 5, 2026
…encode-go-missing-models feat(opencode-go): register 4 missing models from upstream catalog
muhamadgalihsaputra
pushed a commit
to niyatna/NiyatnaRoute
that referenced
this pull request
Sep 27, 2026
…encode-go-missing-models feat(opencode-go): register 4 missing models from upstream catalog
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
opencode-goprovider:qwen3.7-max,mimo-v2-pro,mimo-v2-omni,hy3-preview.https://opencode.ai/zen/go/v1/modelsresponse (16 models). Previously, requests for any of these four IDs failed model resolution before reaching the executor.src/shared/constants/modelSpecs.tsfor the two models whose capabilities are well-documented (qwen3.7-max,mimo-v2-pro).mimo-v2-omnialready had a spec upstream;hy3-previewfalls back to__default__.New Models
qwen3.7-maxqwen3.6-plus(Bailian multimodal — thinking + tools + vision).mimo-v2-promimo-v2-omni(262144 ctx, tools + vision).mimo-v2-omnihy3-preview__default__. Conservative since no public capability docs.All four models route to
/chat/completions(default openai format), matching the existingopencode-goexecutor behavior. No executor/translator changes required.Related Issues
Validation
npm run lintnpm run test:unit—opencode-executor.test.ts32 pass / 0 fail;t31-t33-t34-t38-model-specs.test.ts6 pass / 0 failnpm run test:coverage>= 60%for statements, lines, functions, and branchesTests Added Or Updated
tests/unit/opencode-executor.test.ts— new test caseroutes opencode-go catalog-only models to chat completionscovering all four new models, following the exact pattern of the precedent test added in feat(opencode-go): register 6 missing models from upstream catalog #1510.Coverage Notes
open-sse/config/providerRegistry.ts(data-only registry entries) andsrc/shared/constants/modelSpecs.ts(data-only spec entries). Both are exercised by the new executor test (opencode-executor.test.ts) and the existing model-specs test suite (t31-t33-t34-t38-model-specs.test.ts).Reviewer Notes
feat(opencode-go): register 6 missing models from upstream catalog) added the previous batch via the same gap. This PR follows the same structure (registry + matching executor test) and naming style (namefield in PascalCase to match neighbors likeMiMo-V2.5-Pro,Qwen3.6 Plus).opencode-gocatalog can be verified withcurl https://opencode.ai/zen/go/v1/models— current response includes all 16 models. Before this PR, the registry only listed 12 of them.hy3-preview(Tencent Hunyuan-3 preview) is intentionally added without a model spec — public capability docs are limited, so falling back to__default__is the safest option. A follow-up PR can refine the spec once vision/tools support is confirmed by Tencent docs.Co-Authored-Bytrailer (Hard Rule fix(ci): fix npm publish auth — support vars.NPM_TOKEN #16).