SkillSpector/model_registry.yaml at main · NVIDIA/SkillSpector · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
# Model registry — context window and output token limits.
#
# Point SKILLSPECTOR_MODEL_REGISTRY at this file (or your own) so the tool
# knows each model's token budget.  This is the fallback when the dynamic
# metadata API is unavailable (e.g. open-source deployments).
#
# Format:
#   models:
#     "<model-label>":
#       context_length: <int>          # total context window in tokens (required)
#       max_output_tokens: <int>       # model's max output cap (optional)

models:
  # Stock OpenAI model IDs (for direct api.openai.com or compatible endpoints).
  "gpt-5.2":
    context_length: 400000
    max_output_tokens: 128000

  "gpt-5.3-chat":
    context_length: 128000
    max_output_tokens: 16384

  # Pricing (USD per 1M tokens) is informational only; SkillSpector does not
  # compute cost.  Prompts over 272K input tokens bill at 2x input / 1.5x output.
  # Source: https://developers.openai.com/api/docs/models/<model> (as of 2026-09-21)

  # $4.00 input / $0.40 cached input / $20.00 output (promotional through
  # 2026-11-21).  Input is capped at 922K tokens; the 75% input budget
  # (~787K) stays under it.
  "gpt-5.6-sol":
    context_length: 1050000
    max_output_tokens: 128000

  # $2.00 input / $0.20 cached input / $12.00 output
  "gpt-5.6-terra":
    context_length: 1050000
    max_output_tokens: 128000

  # $0.20 input / $0.02 cached input / $1.20 output
  "gpt-5.6-luna":
    context_length: 1050000
    max_output_tokens: 128000

  # Provider-prefixed IDs for inference gateways that accept them.
  "azure/anthropic/claude-opus-4-5":
    context_length: 200000
    max_output_tokens: 64000

  "azure/anthropic/claude-sonnet-4-6":
    context_length: 1000000
    max_output_tokens: 128000

  "azure/anthropic/claude-opus-4-6":
    context_length: 1000000
    max_output_tokens: 128000

  "openai/openai/gpt-5.2":
    context_length: 400000
    max_output_tokens: 128000

  "openai/openai/gpt-5.3-chat":
    context_length: 128000
    max_output_tokens: 16384

  "gemini-3.5-flash":
    context_length: 1048576
    max_output_tokens: 65536