[model-inventory] Model alias inventory update - 2026-05-05

[github-actions]

Summary

Daily model inventory check for 2026-05-05. Three providers returned data (Anthropic, OpenAI, Gemini); Copilot API returned HTTP 400 error. Several new model families and versions were found that warrant alias and multiplier updates.

• Providers queried: Anthropic, OpenAI, Gemini, Copilot
• Total models found: 257 (9 + 198 + 50 + 0)
• Proposed alias changes: 4
• Multiplier gaps found: 9

Provider Model Counts

Provider	Models Available	Status
anthropic	9	✅ ok
openai	198	✅ ok
gemini	50	✅ ok
copilot	0	❌ HTTP 400

Raw API Fields Discovered

Anthropic (/data[] array):

• id, display_name, created_at, type, max_input_tokens, max_tokens
• capabilities: batch, citations, code_execution, context_management, effort, image_input, pdf_input, structured_outputs, thinking (with types.enabled / types.adaptive)
• Useful fields: max_input_tokens (context window), capabilities.thinking.supported (reasoning model flag), capabilities.image_input.supported (vision flag), capabilities.effort.supported (extended reasoning)

OpenAI (/data[] array):

• id, object, created (Unix timestamp), owned_by
• Very minimal; no token limits, no capability flags in public list endpoint
• owned_by values: openai, openai-internal, system — useful for filtering internal/experimental models

Gemini (/models[] array):

• name (e.g. models/gemini-2.5-pro), version, displayName, description
• inputTokenLimit, outputTokenLimit — useful proxy for model tier
• supportedGenerationMethods — capability flags (generateContent, generateAnswer, generateImage, etc.)
• thinking boolean — flags reasoning/deep-think models
• temperature, topP, topK, maxTemperature

Token Multiplier Analysis

Missing from model_multipliers.json

Model ID	Provider	Inferred Multiplier	Basis
gemini-3-flash-preview	gemini	~0.33	Billing docs: "Gemini 3 Flash" = 0.33 new mult
gemini-3-pro-preview	gemini	~6.0	Billing docs: "Gemini 3 Pro" = 6 new mult
gemini-3-pro-image-preview	gemini	~6.0	Same family as Gemini 3 Pro
gemini-3.1-pro-preview	gemini	~6.0	Billing docs: "Gemini 3.1 Pro" = 6 new mult
gemini-3.1-flash-lite-preview	gemini	~0.1	Flash-Lite tier heuristic
gemini-3.1-flash-image-preview	gemini	~0.33	Flash tier (Gemini 3 Flash)
gemini-2.5-flash-lite	gemini	~0.1	Same tier as gemini-2.5-flash-lite in billing
gemini-2.5-flash-image	gemini	~0.2	Same as gemini-2.5-flash
gpt-5.4-pro	openai	~2.0	Same as gpt-5-pro / gpt-5.5-pro pattern
gpt-5.2-pro	openai	~2.0	Same as gpt-5-pro pattern

Stale entries (no longer returned by any API)

The following entries in model_multipliers.json use the dotted Claude naming convention (e.g. claude-opus-4.5) or reference older Claude 3.x generations that the Anthropic API no longer returns. They may still be served under aliases but are not surfaced in the live model list:

• claude-haiku-4.5, claude-sonnet-4.5, claude-sonnet-4.6, claude-opus-4.5, claude-opus-4.6 (dotted variants — likely alias IDs)
• claude-3-5-haiku, claude-3-haiku, claude-3-sonnet, claude-3-5-sonnet, claude-3-7-sonnet, claude-3-opus, claude-3-5-opus

These should be kept as defensive entries in case the models are still routed, but they are no longer returned by the live Anthropic API.

Inferred vs stored discrepancies

Note: model_multipliers.json uses an internal ET normalization (reference = claude-sonnet-4.5 = 1.0), while GitHub billing multipliers use a separate scale. Direct numeric comparison is not meaningful; the table below highlights relative tier mismatches only.

Model ID	Stored ET Multiplier	GitHub Billing New Mult	Notes
claude-sonnet-4-5	1.0	6	Billing shows significant increase; ET ratio unchanged vs reference
claude-opus-4-5	5.0	15	Opus tier increasing 3×; ET relative ratio may need review
claude-opus-4-6	5.0	27	Very large billing increase; should verify ET multiplier
claude-opus-4-7	5.0	27	Same; needs verification

Proposed Alias Updates

1. Extend gemini-flash to explicitly cover Gemini 3 Flash

What: Add google/gemini-3*flash* pattern to gemini-flash alias.
Why: Gemini 3 Flash (gemini-3-flash-preview, gemini-3.1-flash-*-preview) is a new generation that matches the existing glob but making it explicit aids documentation.

models:
  gemini-flash:
    - "copilot/gemini-*flash*"
    - "google/gemini-*flash*"   # already covers gemini-3-flash-preview ✅

The existing wildcard google/gemini-*flash* already covers all Gemini 3 Flash variants. No change needed — just confirming coverage.

2. Extend gemini-pro to explicitly cover Gemini 3 Pro

What: The existing google/gemini-*pro* glob already covers gemini-3-pro-preview and gemini-3.1-pro-preview.
Why: Confirming the pattern is sufficient; no change needed.

3. New deep-research alias

What: New semantic alias for Gemini deep-research models.
Why: deep-research-* models are a distinct family (specialized research agents with very large context) not covered by any existing alias. They appear in Gemini's model list.

models:
  deep-research:
    - "google/deep-research*"
    - "copilot/deep-research*"

4. New gpt-5-pro alias

What: New alias targeting the GPT-5 Pro tier models.
Why: gpt-5-pro, gpt-5.2-pro, gpt-5.4-pro, gpt-5.5-pro are a distinct high-capability tier above gpt-5 base. These are returned by the OpenAI API and in billing docs but no alias targets them specifically.

models:
  gpt-5-pro:
    - "copilot/gpt-5*pro*"
    - "openai/gpt-5*pro*"

Update large meta-alias to include gpt-5-pro:

models:
  large:
    - "sonnet"
    - "gpt-5-pro"
    - "gpt-5"
    - "gemini-pro"

5. External vendor aliases (Grok, Raptor)

What: GitHub billing docs list "Grok Code Fast 1" and "Raptor mini" as Copilot-served models, suggesting new vendor support.
Why: No IDs for these were found in the live Copilot inventory (HTTP 400 error prevented Copilot API inspection). Once Copilot API access is restored, consider:

models:
  grok-fast:
    - "copilot/grok*fast*"
    - "xai/grok*fast*"

Deferred until Copilot API issue is resolved.

Full Model Lists by Provider

Anthropic (9 models)

• claude-haiku-4-5-20251001
• claude-opus-4-1-20250805
• claude-opus-4-20250514
• claude-opus-4-5-20251101
• claude-opus-4-6
• claude-opus-4-7
• claude-sonnet-4-20250514
• claude-sonnet-4-5-20250929
• claude-sonnet-4-6

Gemini (selected — 50 total)

• deep-research-max-preview-04-2026
• deep-research-preview-04-2026
• deep-research-pro-preview-12-2025
• gemini-2.0-flash, gemini-2.0-flash-001
• gemini-2.0-flash-lite, gemini-2.0-flash-lite-001
• 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-pro
• gemini-3-flash-preview
• gemini-3-pro-image-preview, gemini-3-pro-preview
• gemini-3.1-flash-image-preview, gemini-3.1-flash-lite-preview
• gemini-3.1-pro-preview, gemini-3.1-pro-preview-customtools
• gemini-flash-latest, gemini-flash-lite-latest, gemini-pro-latest
• nano-banana-pro-preview (experimental)
• (plus gemma-3/4, imagen-4, veo-2/3, lyria — non-text models)

OpenAI (selected GPT-5 + reasoning — 198 total)

• gpt-5, gpt-5-2025-08-07, gpt-5-chat-latest
• gpt-5-codex, gpt-5-codex-alpha/beta/mini-alpha/mini-beta
• gpt-5-mini, gpt-5-mini-2025-08-07
• gpt-5-nano, gpt-5-nano-2025-08-07
• gpt-5-pro, gpt-5-pro-2025-10-06
• gpt-5.1, gpt-5.1-2025-11-13, gpt-5.1-chat-latest
• gpt-5.1-codex, gpt-5.1-codex-max, gpt-5.1-codex-mini
• gpt-5.2, gpt-5.2-2025-12-11, gpt-5.2-chat-latest
• gpt-5.2-codex, gpt-5.2-pro, gpt-5.2-pro-2025-12-11
• gpt-5.3-chat-latest, gpt-5.3-codex
• gpt-5.4, gpt-5.4-2026-03-05, gpt-5.4-mini, gpt-5.4-nano, gpt-5.4-pro
• gpt-5.5, gpt-5.5-2026-04-23, gpt-5.5-pro, gpt-5.5-pro-2026-04-23
• o1-pro, o3-2025-04-16, o3-mini, o3-pro, o4-mini, o4-mini-deep-research

Notes

• Copilot API unavailable: The Copilot API returned HTTP 400 for this run. Copilot-prefixed patterns (copilot/*) could not be validated. The analysis relies on direct vendor APIs only.
• Existing wildcard coverage: The current gemini-*flash* and gemini-*pro* globs already cover all Gemini 3.x and 3.1.x models returned by the live API. No urgent pattern changes are needed for these.
• GPT-5 codex variants: The gpt-5-codex alias uses gpt-5*codex* which correctly covers gpt-5.1-codex, gpt-5.2-codex, gpt-5.3-codex etc.
• Billing multiplier scale: The GitHub billing docs multipliers are independent of the internal ET multipliers in model_multipliers.json. The discrepancies table above flags relative tier changes worth reviewing, not direct numeric mismatches.

Generated by Daily Model Inventory Checker · ● 1.2M · ◷

• expires on May 12, 2026, 12:17 AM UTC