Bare anthropic.claude-* Bedrock rows already derived eu.* selectors; also
emit us-gov.* so GovCloud accounts can resolve system inference profiles
without requiring a full partition ARN.
Moved the direct DeepSeek enabled toggle into the thinking-only compat variant and normalized stale cached compat metadata before request encoding.
Fixes#7559
- Introduce `@oh-my-pi/omptype` as a new ArkType-compatible schema validation package featuring a lazy JIT runtime, JSON Schema emission, and compatibility adapters.
- Replace `arktype` across workspace packages and test utilities with `@oh-my-pi/omptype`.
- Add benchmark suites, tests, and documentation for the new validation engine and adapters.
- Update workspace build, test runner, and release configurations to include the new package.
- Account-scoped bearer /v1/models responses now replace the static seed
instead of merging, so models disabled for the account are not selectable.
- Extended the catalog regression to run a real online refresh and assert
the static seeds are pruned to the fetched IDs.
Applied GitHub Copilot's discovered default token-price tier to base models while preserving the provider fallback for unreported cache-write costs. Added regression coverage for GPT-5.6 Luna base and long-context pricing.
Fixes#7471
Dynamically discovered deepseek-v4* models (e.g. deepseek-v4-flash-0731)
now receive reasoning: true and [high, max] thinking efforts via prefix
matching in the discovery mapper.
Filtered text-embedding model IDs from authoritative Alibaba Token Plan chat discovery.
Covered text-embedding-v4 alongside the existing media-only discovery fixtures.
Fixes#7391
Replaced the stale static discovery allowlist with explicit filters for media-only model families.
Covered newly advertised DeepSeek, Kimi, and MiniMax chat models while retaining image, audio, and video exclusions.
Fixes#7391
Use isOpenCodeHost (provider id + baseUrl markers) instead of the built-in
provider-id check so DeepSeek reasoning models declared under a custom provider
id pointed at the OpenCode gateway URL also downgrade a forced tool_choice to
auto.
Fixes#7315
Limit the DeepSeek forced tool-choice downgrade to the OpenCode Zen and Go
gateways whose default thinking mode rejects named selectors. Preserve forced
tool selection on NVIDIA and other gateways that can disable thinking for the
request.
Add regression coverage for both the affected OpenCode path and an unaffected
NVIDIA DeepSeek route.
Fixes#7315
DeepSeek reasoning models on the OpenCode Zen/Go gateways 400 with
"Thinking mode does not support this tool_choice" when a specific tool is
forced. The compat descriptor already drops reasoning_effort via
disableReasoningOnToolChoice, but that does not turn off the gateway's
default thinking mode, so the forced named selector still trips DeepSeek's
thinking+tool_choice guard.
Mark forced tool_choice unsupported for DeepSeek reasoning models so
buildParams downgrades the selector to "auto" while keeping the tool
advertised, mirroring the Anthropic and direct-DeepSeek paths.
Fixes#7315
- Claude ids now alias to google-vertex suffixed entries (claude-opus-4-6@default etc.) so Antigravity follows Google's price if it diverges from Anthropic's list price; plain-id anthropic lookup remains as dangling-alias fallback.
- Regenerated models.json via gen:models.
- Added `applyAntigravityPricingFallback` to back-fill unpriced `google-antigravity` models using first-party list prices from `google` and `anthropic`.
- Mapped Antigravity model IDs to preview-id aliases (e.g., `gemini-3.1-pro` to `gemini-3.1-pro-preview`) for correct pricing resolution.
- Added test coverage in `packages/catalog/test/generated-policies.test.ts` verifying fallback pricing and preservation of billable costs.
- Parsed OpenRouter reasoning effort ladders and defaults during discovery.
- Preserved explicit thinking metadata from models.yml patches.
- Regenerated the catalog and covered both regression paths.
Fixes#7307
- Rewrote fetchCodexDiscoveryModels to resolve every stored openai-codex
OAuth account via getOAuthAccesses and reuse openaiCodexModelManagerOptions'
tested union/fail-closed path, so a single narrow account can no longer
authoritatively wipe sibling-account models from the bundle.
- Restored openai-codex gpt-5.4, gpt-5.6-sol, and gpt-5.3-codex-spark bundle
entries (and gpt-5.5's contextPromotionTarget) dropped by the previous
single-account regen.
- Updated the live Codex image tool-result tests off the retired
gpt-5.2-codex id to gpt-5.5.
Fixes#6265
Ollama Cloud's deepseek-v4-pro and deepseek-v4-flash deployments reject any
output budget above 65536 with HTTP 400, despite advertising a 1M context /
384K output (ollama/ollama#16890). Ollama's /api/show never reports this cap,
so the catalog left the base models at the full context window and the dated
tag deepseek-v4-flash:0731 at a stale 8192 fallback. Pin these ids (base plus
tag variants) to min(contextWindow, 65536) at both runtime discovery and
generation; other cloud models keep their discovered limits.
Fixes#7266
anthropic-messages requests dispatched through the direct Anthropic provider
path ignored the catalog compat idle timeout. Slow relayed endpoints therefore
remained limited by the default inter-event watchdog even when the model
configured a wider or disabled timeout.
Add streamIdleTimeoutMs to AnthropicCompat, preserve it through
buildAnthropicCompat, and pass it as the fallback to getStreamIdleTimeoutMs in
streamAnthropic. A value of 0 disables only the inter-event watchdog; the
first-event watchdog is unchanged.
Resolution order matches the existing timeout helper contract:
caller option
> PI_STREAM_IDLE_TIMEOUT_MS
> PI_OPENAI_STREAM_IDLE_TIMEOUT_MS
> model compat
> 300s default
The pi-native transport retains its existing transport-level timeout policy.
Tested: anthropic stream timeout suite, env-contaminated regression cases,
models config validation, and bun check.
- Implement the ai& provider registry entry with API-key authentication and login support.
- Add model descriptors, static model seeding, and openai-compatible model discovery for the ai& provider.
- Update the model catalog with ai& provider models, pricing, and updated provider model names.
- Add unit tests for the ai& provider environment resolution, metadata, and dynamic model mapping.
- Migrated model catalog fetching and documentation references from the models.dev API to the stencil.so well-known models endpoint.
- Added support for zstd decompression and session-based caching with ETag conditional requests and stale fallback handling.
- Updated test suites, mock URLs, and constants across catalog and coding-agent packages to target stencil.so.
GMI's launch post documents only $0.14/$0.28 per 1M for DeepSeek-V4-Flash;
the $0.028 cache-read figure was back-derived from OpenRouter's discounted
route and has no direct api.gmi-serving.com source. Keep cacheRead at 0
until GMI confirms cached-token billing, so usage cost reporting is not
overstated.
- Set deepseek-ai/DeepSeek-V4-Flash cost to GMI's direct api.gmi-serving.com
tariff ($0.14/$0.28 per 1M, cache read $0.028) per GMI's DeepSeek V4
launch post, instead of values inherited from other providers.
- Documented that mapWithBundledReference keeps the seed's cost/reasoning/
thinking at runtime: /v1/models discovery only overrides the model ID set
and context/max-token limits, so the seed must carry real GMI values.
- Regenerated models.json.
- Added GMI_CLOUD_STATIC_MODELS seed wired into gen:models so a fresh
install resolves deepseek-ai/DeepSeek-V4-Flash synchronously at boot,
before async /v1/models discovery fires; live discovery stays
authoritative and replaces the seed.
- Regenerated models.json with the seeded gmi-cloud slice.
- Dropped the GeneratedProvider cast now that gmi-cloud is bundled.
- Moved gmi-cloud to the end of the /login provider order.
- Moved changelog entries from released 17.0.3 sections to Unreleased.
- Added a regression test asserting the seed covers the descriptor's
defaultModel.
Google AI Studio's OpenAI-compatible endpoint
(generativelanguage.googleapis.com/v1beta/openai/chat/completions)
implements a subset of the chat-completions schema and rejects the
`store` field with HTTP 400:
Invalid JSON payload received. Unknown name "store": Cannot find field.
OMP unconditionally emits `store: false` for any openai-completions
model whose resolved compat has supportsStore: true, so every request
to a custom provider pointed at this host (a common wiring for the
`vision` role, e.g. gemini-2.5-flash) failed before the first token.
Add a googleAistudio host class (URL-marker only, like chutes, since
users wire this host under arbitrary provider ids in models.yml) and
include it in the isNonStandard set so supportsStore resolves false.
Same failure shape as openclaw/openclaw#22704.
Detection is URL-based: verified against a captured 400 request dump
from generativelanguage.googleapis.com/v1beta/openai/chat/completions.
Add the gmi-cloud chat-model provider, an OpenAI-compatible inference
gateway at https://api.gmi-serving.com/v1. Wired per the adding-a-provider
contract: a createSimpleOpenAICompletionsOptions helper and catalog entry
in pi-catalog, plus a registry definition with an API-key paste login in
pi-ai. Dynamic model discovery via /v1/models; auth falls back to
GMI_API_KEY. defaultModel is deepseek-ai/DeepSeek-V4-Flash.
The providerId is cast to GeneratedProvider until the next gen:models run
seeds gmi-cloud into models.json (same pattern as xai-oauth/vllm).
Codex P2 on b88869dc6 (reproduced live): the mapper's `reasoning: false`
for a `none`-only route only held for the raw fetcher. The production path
normalizes through `createModelManager`, where `mergeDynamicModel` merged
the dynamic row over the bundled reference with
`existingModel.reasoning || dynamicModel.reasoning` — so a stale bundled
`reasoning: true` (e.g. `hf:zai-org/GLM-5.2`) won again and `buildModel`
fabricated a `[minimal,low,medium,high,xhigh]` ladder for a route that
advertised only the `none` off-state.
`mergeDynamicModel` now treats Synthetic's discovered `reasoning` as
authoritative (both-sides-synthetic guard inside the per-id merge),
mirroring the existing Copilot `dynamicInputAuthoritative` precedent in
the same function. This loses nothing: the Synthetic mapper already folds
the reference's reasoning vote into the dynamic row whenever the wire is
silent on reasoning, so the reference still gets its say — at the mapper,
where the wire vocabulary can be consulted, instead of via a blind OR at
the merge.
The new regression test drives the full production path
(`createModelManager.refresh("online")`) rather than calling the fetcher
and `buildModel` directly, closing the gap Codex called out.
Codex P2 on b88869dc6: the reasoning gate required more than one ladder
entry, so a route advertising exactly one named tier (e.g.
`reasoning_parameters.efforts: ["high"]`) was marked non-reasoning and its
only wire-accepted `reasoning_effort` was hidden and dropped. The
off-switch is a vocabulary with no named tiers (`["none"]` or unrecognized
values alone), not a one-tier ladder — gate on the named-tier count
instead. Adds a single-tier regression test asserted through buildModel.
Addresses two review items on the Synthetic capability fix:
- Codex P2: an advertised `reasoning_parameters.efforts` list is now
authoritative over the bundled reference's `reasoning` flag. Previously a
`hf:*` reference with `reasoning: true` re-armed the effort dial even when
the wire advertised only the `none` off-state, so the raw spec carried a
misleading `reasoning: true` plus a one-stop minimal ladder. When the wire
names tiers, the reference no longer gets a vote.
- roboomp should-fix: a present-but-empty `supported_features: []` was
treated like absent metadata and left `supportsTools` unset, which the
request layer reads as tool-capable. A present array (empty included) with
no `tools` entry now marks the route tool-less, while a reference that
already vouched for tools still wins over a merely-incomplete wire list.
Tests: stale-reference override is asserted through the built-model path
(`buildModel`), and a new fixture covers the empty-features contract.
7 pass; full packages/catalog suite 525 pass, 0 fail.
Code review on b2880217e found three edge-shape leaks in the new mapper:
- A route advertising efforts: ["none"] (or only unrecognized tiers) set
reasoning: true with thinking: undefined, so resolveModelThinking fell
through to identity inference and fabricated an unadvertised
low/medium/high/xhigh ladder for the request layer. None-only routes
now map onto minimal -> none and report non-reasoning; unrecognized-only
vocabularies keep the reference reasoning flag but never grow tiers the
wire did not name.
- The same undefined-thinking path let a stale bundled reference ladder
(e.g. hf:zai-org/GLM-5.2's baked tiers) survive past the wire
vocabulary. Wire efforts now always override the reference ladder.
- supportsTools: false was written from a populated but tool-less
supported_features list with no reference OR-fallback, unlike
reasoning/input, so an incomplete wire list could hard-disable tools
on a route the reference marked tool-capable. The strip now also
honors an explicit reference supportsTools: false and only downgrades
when neither source vouches for tools.
Tests: none-only vocabulary, stale-reference override, and the discovery
request's Authorization header. 6 pass; full packages/catalog suite 524
pass, 0 fail.
The Synthetic discovery mapper looked for `supports_reasoning`,
`supports_vision` and `max_tokens`. Synthetic's /openai/v1/models sends
none of those: capabilities arrive in `supported_features`, the accepted
`reasoning_effort` vocabulary in `reasoning_parameters.efforts`,
modalities in `input_modalities`, the output cap in `max_output_length`,
and per-token prices in `pricing` as `$`-prefixed strings.
Every capability therefore fell back to the bundled reference, so any
route without one — the `syn:*` router aliases, newly added routes such
as `hf:moonshotai/Kimi-K3` — resolved to `reasoning: false`, text-only,
zero cost, `maxTokens: 8192`. `reasoning: false` makes
`getSupportedEfforts` return `[]`, which hides the thinking selector and
silently drops a `provider/model:high` suffix, so those models never
receive `reasoning_effort` at all; the 8k cap is low enough that verbose
models stop on `length` every turn, which reads as an incomplete
response and triggers recovery compaction at ~9% context.
Read the fields Synthetic actually sends and derive the effort ladder
from the per-model wire vocabulary. `none` is the router's thinking-off
state rather than a user tier, so it backs `minimal` through the wire
map, mirroring the existing Fireworks `minimal -> none` mapping.
Verified live against api.synthetic.new: `reasoning_effort` is accepted
on every route (`none` yields no reasoning, `high` < `max`), and the
`syn:*:vision` aliases do accept image input.
Pete Samwel