[model-inventory] Model alias inventory - no changes needed - 2026-08-25

[github-actions]

Summary

Analyzed live model inventories from OpenAI, Anthropic, Gemini, Copilot SDK, and Copilot
(reflect), and cross-checked against pkg/workflow/data/model_aliases.json and
pkg/cli/data/models.json/actions/setup/js/models.json. No alias or pricing updates are
needed. OpenAI and Anthropic provider API keys were not configured for this run (HTTP 401 /
not set), so those two providers' live coverage checks fall back to the pre-fetched
models.json snapshot in the artifacts; Copilot (reflect + SDK) and Gemini were fully live.
The Copilot Docs pricing page could not be scraped (playwright-cli not available in this
environment), so billing-table cross-check was skipped for this run — SDK and reflect pricing
were used instead and are consistent with models.json.

• Providers queried: OpenAI (no key), Anthropic (9 models, snapshot), Gemini (50 models, live),
  Copilot SDK (24 models, live), Copilot reflect (35 models, live)
• Total models found: 118
• Proposed alias changes: 0
• Pricing gaps found: 0 (all discrepancies are intentionally-excluded legacy IDs or
  historical/retired entries)

Provider Model Counts

Provider	Models Available	Status
openai	0	❌ no API key (HTTP 401)
anthropic	9	✅ ok
gemini	50	✅ ok
copilot-sdk	24 (23 priced + auto)	✅ ok
copilot (reflect)	35	✅ ok
models.dev/api.json	(secondary cross-check catalog, not scored)	✅ available

Raw API Fields Discovered

• Anthropic (raw.json): id, display_name, created_at, type — no pricing/capability
  fields in the public models-list response; family grouping (haiku/opus/sonnet) is
  inferable from display_name only.
• Gemini (raw.json): id, display_name, description, input_token_limit,
  output_token_limit, supported_generation_methods, version — no explicit billing/tier
  field; context/output limits are usable as a rough capability-tier proxy.
• Copilot SDK (copilot-sdk-models/models.json): rich billing.tokenPrices (input/output/
  cache read/cache write, contextMax, longContext tier), capabilities.limits (context/
  output/vision), capabilities.supports (tool_calls, vision, reasoning_effort, adaptive
  thinking), modelPickerCategory, modelPickerPriceCategory, policy.state. This is the
  richest and most authoritative pricing source available.
• Copilot reflect (reflect.json → model_metadata[].pricing): per-model default/
  long_context tiers with input/output/cachedInput/cacheWrite (per-million-token
  units), matching SDK values 1:1 after unit conversion (reflect uses $/M tokens, models.json
  stores $/token).
• OpenAI: unavailable this run (no key configured).

All of the above fields worth retaining are already captured by the existing enriched
models.json/model_aliases.json maintenance flow; no new fields need to be added to the
inventory schema.

models.json Pricing Analysis

Missing from models.json

None. All live Copilot SDK model IDs are present in pkg/cli/data/models.json /
actions/setup/js/models.json. The Copilot reflect provider list also included legacy/utility
IDs (gpt-3.5-turbo*, gpt-4, gpt-4-0613, gpt-4-o-preview, gpt-4.1-2025-04-14,
gpt-41-copilot, gpt-4o-2024-*, gpt-4o-mini-2024-07-18, text-embedding-*,
trajectory-compaction) that are either embedding/utility models, dated snapshot aliases of
already-covered families, or deprecated per this task's exclusion list — none warrant new
top-level pricing entries.

Historical entries not currently returned

These models.json github-copilot entries were not present in this run's live SDK/reflect
inventories and are retained as historical records (no removal proposed):

claude-fable-5, claude-opus-4.5, claude-opus-4.6, claude-sonnet-4, gemini-2.5-pro,
gemini-3-flash-preview, gpt-5.2, gpt-5.2-codex, gpt-5.4-nano (intentionally excluded
per task instructions), kimi-k2.7-code, kimi-k3, raptor-mini.

Inferred vs stored pricing discrepancies

None found. Spot-checked claude-sonnet-5, claude-opus-5, gpt-5.5, gpt-5.4-mini — reflect
pricing.default ($/M tokens) and SDK billing.tokenPrices agree with models.json cost
($/token) after the 1e6 unit conversion in every case checked.

Proposed Alias Updates

None. Every live model ID returned by Copilot (reflect + SDK), Anthropic, and Gemini in this run
matched at least one existing glob pattern in pkg/workflow/data/model_aliases.json (verified
programmatically with fnmatch against all vendor/pattern* entries). Existing aliases already
cover the gemini-3.x/gpt-5.x/claude-*-5/grok-4.x/mai-code/kimi generations, and
task-oriented aliases (summarization, coding, reasoning, vision, computer-use) are
already present.

Full Model Lists by Provider

Anthropic (9): claude-haiku-4-5-20251001, claude-opus-4-5-20251101, claude-opus-4-6,
claude-opus-4-7, claude-opus-4-8, claude-opus-5, claude-sonnet-4-5-20250929, claude-sonnet-4-6,
claude-sonnet-5

Gemini (50): antigravity-preview-05-2026, aqa, deep-research-max-preview-04-2026,
deep-research-preview-04-2026, deep-research-pro-preview-12-2025,
gemini-2.5-computer-use-preview-10-2025, gemini-2.5-flash, gemini-2.5-flash-image,
gemini-2.5-flash-lite, gemini-2.5-flash-native-audio-latest,
gemini-2.5-flash-native-audio-preview-09-2025, gemini-2.5-flash-native-audio-preview-12-2025,
gemini-2.5-flash-preview-tts, gemini-2.5-pro, gemini-2.5-pro-preview-tts,
gemini-3-flash-preview, gemini-3-pro-image, gemini-3-pro-image-preview,
gemini-3.1-flash-image, gemini-3.1-flash-image-preview, gemini-3.1-flash-lite,
gemini-3.1-flash-lite-image, gemini-3.1-flash-lite-preview, gemini-3.1-flash-live-preview,
gemini-3.1-flash-tts-preview, gemini-3.1-pro-preview, gemini-3.1-pro-preview-customtools,
gemini-3.5-flash, gemini-3.5-flash-lite, gemini-3.5-live-translate-preview, gemini-3.6-flash,
gemini-3.7-flash, gemini-embedding-001, gemini-embedding-2, gemini-embedding-2-preview,
gemini-flash-latest, gemini-flash-lite-latest, gemini-omni-flash-preview, gemini-pro-latest,
gemini-robotics-er-1.6-preview, gemini-robotics-er-2-preview,
gemini-robotics-er-2-streaming-preview, gemma-4-26b-a4b-it, gemma-4-31b-it,
lyria-3-clip-preview, lyria-3-pro-preview, nano-banana-pro-preview, veo-3.1-fast-generate-preview,
veo-3.1-generate-preview, veo-3.1-lite-generate-preview

Copilot SDK (24): auto, claude-sonnet-5, claude-opus-5, claude-opus-4.8, claude-opus-4.7,
claude-sonnet-4.6, claude-sonnet-4.5, claude-haiku-4.5, gpt-5.6-sol, gpt-5.6-terra,
gpt-5.6-luna, gpt-5.5, gpt-5.4, gpt-5.4-mini, gpt-5.3-codex, gpt-5-mini,
mai-code-1-flash-picker, gemini-3.7-flash, gemini-3.6-flash, gemini-3.5-flash,
gemini-3.1-pro-preview, grok-4.5, grok-4.6, mai-code-1.1-flash

Copilot reflect (35): claude-haiku-4.5, claude-opus-4.7, claude-opus-4.8, claude-opus-5,
claude-sonnet-4.5, claude-sonnet-4.6, claude-sonnet-5, gpt-3.5-turbo, gpt-3.5-turbo-0613, gpt-4,
gpt-4-0613, gpt-4-o-preview, gpt-4.1, gpt-4.1-2025-04-14, gpt-41-copilot, gpt-4o,
gpt-4o-2024-05-13, gpt-4o-2024-08-06, gpt-4o-2024-11-20, gpt-4o-mini, gpt-4o-mini-2024-07-18,
gpt-5-mini, gpt-5.3-codex, gpt-5.4, gpt-5.4-mini, gpt-5.5, gpt-5.6-luna, gpt-5.6-sol,
gpt-5.6-terra, grok-4.5, grok-4.6, mai-code-1-flash-picker, mai-code-1.1-flash,
text-embedding-3-small, text-embedding-3-small-inference, text-embedding-ada-002,
trajectory-compaction

Notes

• OpenAI credentials were not configured for this run; treat OpenAI coverage as unverified
  until a future run has a valid key.
• playwright-cli was not available in this execution environment, so the Copilot docs pricing
  table cross-check (Step 2.5) was skipped; SDK + reflect pricing agreement was used as the
  primary validation and is sufficient (both sources matched models.json exactly).
• gpt-4o-mini, gpt-4.1, gpt-4o, and gpt-5.4-nano were treated as intentionally
  deprecated per task instructions and excluded from all gap/discrepancy tables.
• Historical entries listed above were left untouched per instructions (no automatic removals
  proposed).

Generated by 📦 Daily Model Inventory Checker · copilot · auto · 50.6 AIC · ⌖ 6.16 AIC · ⊞ 13.2K · ◷

• expires on Aug 31, 2026, 4:04 PM UTC-08:00

[github-actions]

This issue is being closed as outdated. A newer issue has been created: #55900

View newer issue

---

This action was performed automatically by the Daily Model Inventory Checker workflow.