Clinical Benchmarks
Menu

Claude Haiku 4.5

Released 15 Oct 2025$1 input, $5 output per million tokens200K contextproprietaryAlso written as claude-haiku-4-5, Claude Haiku 4.5 (Thinking), claude-haiku-4-5-20251001-thinking, anthropic/claude-haiku-4-5-20251001-thinking

5 results on 5 of 13 boards, latest measured Oct 2026. Compare with other models

Clinical Benchmarks Index23.1Rank 98 of 1494 of 10 boardsAverage × multiplier 28.3 × 0.816 = 23.1This site’s own calculation. Working How it is computed

Results

Each line sits on its own board's scale. Tall tick: the board leader.

Clinical reasoning and knowledge

  1. HealthBench Professional
    Anthropic; extended thinking (Haiku 4.5 has no effort setting); Opus 4.8 grader; safety classifiers enabled; length-adjusted; measured for the Haiku 5.5 card
    0.322
    Rank 35 of 35Leader GPT-6 Astra (Anthropic run) 0.703
    Vendor-reported
    Measured Oct 2026

Documentation and coding

  1. MedScribe (Vals AI)
    model ID anthropic/claude-haiku-4-5-20251001-thinking; temperature=1; max_output_tokens=30000
    85.23
    Rank 28 of 105Leader Claude Opus 5.5 91.43
    Official leaderboard
    Measured Sep 2026
  2. MedCode (Vals AI)
    model ID anthropic/claude-haiku-4-5-20251001-thinking; temperature=1; max_output_tokens=30000
    32.68
    Rank 84 of 103Leader Claude Opus 5 63.57
    Official leaderboard
    Measured Sep 2026

EHR and workflow agents

  1. PhysicianBench
    Anthropic harness; 100 tasks; pass@1; extended thinking (no effort setting); Opus 5 rubric grader; safety classifiers enabled; measured for the Haiku 5.5 card
    1.4
    Rank 28 of 28 here, 12 models on the boardLeader Claude Opus 5.5 (max) 68.4
    Vendor-reported
    Measured Oct 2026
  2. CHI-Bench
    Listed as claude-code + claude-haiku-4-5
    All Domains pass@1; claude-code harness
    6.2
    Rank 36 of 43 here, 45 models on the boardLeader erius + claude-opus-5 54.7
    Official leaderboard
    Measured Apr 2026

Index working

The Clinical Benchmarks Index is this site's own calculation: a weighted average of the board scores times a multiplier for the boards behind it. How it is computed

  • HealthBench Professional0.32232.2×0.743Anthropic; extended thinking (Haiku 4.5 has no effort setting); Opus 4.8 grader; safety classifiers enabled; length-adjusted; measured for the Haiku 5.5 card
  • MedScribe (Vals AI)85.2385.2×0.25, near its ceilingNear its ceiling: leader 91.4, weight 0.25model ID anthropic/claude-haiku-4-5-20251001-thinking; temperature=1; max_output_tokens=30000
  • MedCode (Vals AI)32.6832.7×0.911model ID anthropic/claude-haiku-4-5-20251001-thinking; temperature=1; max_output_tokens=30000
  • PhysicianBench1.41.4×0.79Anthropic harness; 100 tasks; pass@1; extended thinking (no effort setting); Opus 5 rubric grader; safety classifiers enabled; measured for the Haiku 5.5 card
Weighted average of 4 boards
28.3
Multiplier for 4 of 10 boards
× 0.816
Index
28.3 × 0.816 = 23.1

Score: the result on the board’s own 0 to 100 scale. Dot: this model; tick: the board’s leader. Weight: how much the board counts in the weighted average; bold where the board is near its ceiling and counts for less. Each result links to its quote and page; the line under it is the setting it was measured with.

Sources

The quote, page and document for each of the 5 results
  1. 35HealthBench Professional extended thinking 0.322
    Printed as 32.2Vendor-reported, measured Oct 2026Configuration: Anthropic; extended thinking (Haiku 4.5 has no effort setting); Opus 4.8 grader; safety classifiers enabled; length-adjusted; measured for the Haiku 5.5 card
    Claude Haiku 5.5 System Card system card, Anthropic, 7 Oct 2026. p. 111, Table 8.1.A, row HealthBench Professional (Length-adjusted), column Haiku 4.5
    Evaluation | Haiku 5.5 | Haiku 4.5 | Sonnet 5.5 | GPT-6 Luna HealthBench Professional (Length-adjusted) | 64.8 | 32.2 | 69.2 | –
    Every result from this document
  2. 28MedScribe (Vals AI) claude-haiku-4-5-20251001-thinking 85.23
    Printed as 85.23%Official leaderboard, measured Sep 2026Configuration: model ID anthropic/claude-haiku-4-5-20251001-thinking; temperature=1; max_output_tokens=30000
    Vals AI MedScribe leaderboard official leaderboard, Vals AI, 3 Sep 2026. Vals AI MedScribe leaderboard, View: All Models, Task: Overall, row 29 of 105 (Claude Haiku 4.5 (Thinking)), Accuracy column; Updated 9/29/2026. Rendered BenchmarkView table; configuration from embedded astro-island BenchmarkView props, benchmarkView.default.tasks.overall["anthropic/claude-haiku-4-5-20251001-thinking"].
    29 | Claude Haiku 4.5 (Thinking) | 85.23%±1.90 | $1/$5 | 66.20s
    Every result from this document
  3. 84MedCode (Vals AI) claude-haiku-4-5-20251001-thinking 32.68
    Printed as 32.68%Official leaderboard, measured Sep 2026Configuration: model ID anthropic/claude-haiku-4-5-20251001-thinking; temperature=1; max_output_tokens=30000
    Vals AI MedCode leaderboard official leaderboard, Vals AI, 29 Sep 2026. MedCode leaderboard, Overall task, All Models expanded, rank 84 of 103; Accuracy column; board Updated 9/29/2026. Rendered table row; configuration from embedded BenchmarkView props benchmarkView.default.tasks.overall["anthropic/claude-haiku-4-5-20251001-thinking"].
    84 | Claude Haiku 4.5 (Thinking) | 32.68%±2.00 | $1/$5 | 34.29s
    Every result from this document
  4. 28PhysicianBench Anthropic harness 1.4
    Printed as 1.4%Vendor-reported, measured Oct 2026Configuration: Anthropic harness; 100 tasks; pass@1; extended thinking (no effort setting); Opus 5 rubric grader; safety classifiers enabled; measured for the Haiku 5.5 card
    Claude Haiku 5.5 System Card system card, Anthropic, 7 Oct 2026. p. 132, Figure 8.11.2.A PhysicianBench pass rate, Claude Haiku 4.5 bar label (visually read printed label; not repeated in the text)
    Claude Haiku 4.5 | 1.4%
    Every result from this document
  5. 36CHI-Bench All Domains pass@1 6.2
    Printed as 6.2%Official leaderboard, measured Apr 2026Configuration: All Domains pass@1; claude-code harness
    CHI-Bench leaderboard (actAVA) official leaderboard, actAVA, 12 Aug 2026. CHI-Bench v1.0.0, All Domains, rank 37, Agent/Model and Accuracy columns; PA/UM/CM follow; board last updated 2026-08-12; board Date column 2026-05-01; accessed 2026-09-30. Linked submission provenance: run 2026-04-29T19:36:29Z to 2026-04-30T19:34:41Z; submitted_at 2026-04-30T19:34:41Z.
    37 | claude-code | claude-haiku-4-5 | Proprietary | 6.2% | 0.0% | 14.7% | 4.0% | 2026-05-01
    Every result from this document

Other Anthropic models: Claude 3.7 Sonnet, Claude Fable 5, Claude Fable 5.1, Claude Haiku 5.5, Claude Opus 4.1, Claude Opus 4.5, Claude Opus 4.6, Claude Opus 4.7, Claude Opus 4.8, Claude Opus 5, Claude Opus 5.5, Claude Sonnet 4, Claude Sonnet 4.5, Claude Sonnet 4.6, Claude Sonnet 5, Claude Sonnet 5.5