Clinical Benchmarks

Thinking Machines

Thinking Machines models with results here, each shown on its own board.

Models
2
Results
5
Boards
3 of 13

thinkingmachines.ai

Models

ModelKindReleasedBoards
Inkling SmallModel30 Jul 20262
InklingModel15 Jul 20263

Results

One line per result. Tall tick: the board leader.

Documentation and coding

  1. MedScribe (Vals AI)
    model ID thinkingmachines/inkling; temperature=1; top_p=1; max_output_tokens=30000; reasoning_effort=0.99
    85.41
    Rank 25 of 105Leader Claude Opus 5.5 91.43
    Official leaderboard
    Measured Sep 2026
  2. MedScribe (Vals AI)
    model ID thinkingmachines/inkling-small; temperature=1; top_p=1; max_output_tokens=30000; reasoning_effort=0.99
    84.11
    Rank 35 of 105Leader Claude Opus 5.5 91.43
    Official leaderboard
    Measured Sep 2026
  3. MedCode (Vals AI)
    model ID thinkingmachines/inkling; reasoning_effort=0.99; temperature=1; top_p=1; max_output_tokens=30000
    41.19
    Rank 53 of 103Leader Claude Opus 5 63.57
    Official leaderboard
    Measured Sep 2026
  4. MedCode (Vals AI)
    model ID thinkingmachines/inkling-small; reasoning_effort=0.99; temperature=1; top_p=1; max_output_tokens=30000
    37.89
    Rank 71 of 103Leader Claude Opus 5 63.57
    Official leaderboard
    Measured Sep 2026

EHR and workflow agents

  1. CHI-Bench
    All Domains pass@1; openai-agents; Tinker 256K, high reasoning, 50-turn limit
    8.0
    Rank 35 of 43 here, 45 models on the boardLeader erius + claude-opus-5 54.7
    Official leaderboard
    Measured Jul 2026