HealthBench Professional

DDeepSeek V4.1 Flash on HealthBench Professional

Rank 33 of 43, updated November 2, 2026.

DeepSeek V4.1 Flash scores 0.507 on HealthBench Professional, rank 33 of 43 models on the board. HealthBench Professional scores models on 525 tasks drawn from real clinician conversations, graded criterion by criterion against physician-written rubrics on a 0 to 1 scale.

33
rank of 43
0.507
length-adjusted score
0.529
score before the length adjustment
0.196
gap to the leader, GPT-6 Astra (Anthropic run)
1M
context window, tokens
$0.30 / $1.20
input / output price per 1M tokens

Where it sits on the board

The models ranked around DeepSeek V4.1 Flash, by length-adjusted score. Differences are against DeepSeek V4.1 Flash.

RankModelLength-adjustedBefore the adjustmentDifference
30Anthropic logoClaude Opus 4.70.519–+0.012
31OpenAI logoGPT-5.50.5180.572+0.011
32Google logoGemini 3.8 Flash0.5170.590+0.010
33DDeepSeek V4.1 Flash0.5070.529–
34SpaceX AI logoGrok 4.60.485–-0.022
35OpenAI logoGPT-5.40.4810.519-0.026
36OpenAI logoGPT-50.4620.510-0.045

Result and API facts

Rank33 of 43
Length-adjusted score0.507
Score before the length adjustment0.529
LabDeepSeek
Context window1M tokens
API price per 1M tokens$0.30 in / $1.20 out
Licenseopen
SourceHealthBench Professional Benchmark | Artificial Analysis (independent leaderboard)
Released2026-09-10

Source of this score

Read from HealthBench Professional Benchmark | Artificial Analysis (independent leaderboard, Artificial Analysis, 2026-11-02). Independent run. Confidence: partial. Configuration: Artificial Analysis independent run, max reasoning; GPT-5.4 (low) judge, 4 repeats per task; length-adjusted (52.9 raw, 2,760 mean response chars).

Score section, Length-Adjusted Score chart, DeepSeek V4.1 Flash (max) bar label; raw score 52.9% on the Raw Score tabfull entry on the sources page
HealthBench Professional: Length-Adjusted Score | DeepSeek V4.1 Flash (max) | 50.7%

Position in the field

The gap to the leader, GPT-6 Astra (Anthropic run) at 0.703, is 0.196. Directly above sits Gemini 3.8 Flash at 0.517. Directly below sits Grok 4.6 at 0.485. The scores on this page are compiled from published documents rather than from one controlled run, so a small gap between 2 models can reflect a difference in grader version, reasoning effort, or deployment setting as well as a difference in capability.

What does DeepSeek V4.1 Flash score on HealthBench Professional?

DeepSeek V4.1 Flash scores 0.507 on HealthBench Professional, which places it at rank 33 of 43 models on the board as of November 2, 2026. The number was read from HealthBench Professional Benchmark | Artificial Analysis, listed on the sources page.

How much does DeepSeek V4.1 Flash cost to run?

DeepSeek V4.1 Flash is priced at $0.30 per million input tokens and $1.20 per million output tokens through DeepSeek's API.

Pairings with a dedicated comparison page are linked; every other difference is in the score-difference matrix.

  • DeepSeek V4.1 Flash vs GPT-6 Astra (Anthropic run)
    0.507 vs 0.703, GPT-6 Astra (Anthropic run) by 0.196
  • DeepSeek V4.1 Flash vs Claude Sonnet 5.5
    0.507 vs 0.692, Claude Sonnet 5.5 by 0.185
  • DeepSeek V4.1 Flash vs Claude Fable 5
    0.507 vs 0.660, Claude Fable 5 by 0.153
  • DeepSeek V4.1 Flash vs Claude Opus 5.5
    0.507 vs 0.656, Claude Opus 5.5 by 0.149
  • DeepSeek V4.1 Flash vs GPT-6 Luna (Anthropic run)
    0.507 vs 0.655, GPT-6 Luna (Anthropic run) by 0.148
  • DeepSeek V4.1 Flash vs Claude Haiku 5.5
    0.507 vs 0.648, Claude Haiku 5.5 by 0.141
  • DeepSeek V4.1 Flash vs GPT-6 Astra
    0.507 vs 0.647, GPT-6 Astra by 0.140
  • DeepSeek V4.1 Flash vs GPT-6.1 Sol
    0.507 vs 0.642, GPT-6.1 Sol by 0.135
  • DeepSeek V4.1 Flash vs Claude Fable 5 (September card)
    0.507 vs 0.633, Claude Fable 5 (September card) by 0.126
  • DeepSeek V4.1 Flash vs GPT-6.1 Terra
    0.507 vs 0.627, GPT-6.1 Terra by 0.120
  • DeepSeek V4.1 Flash vs Muse Wonder Full
    0.507 vs 0.623, Muse Wonder Full by 0.116
  • DeepSeek V4.1 Flash vs Claude Fable 5.1
    0.507 vs 0.621, Claude Fable 5.1 by 0.114
  • DeepSeek V4.1 Flash vs GPT-6 Luna
    0.507 vs 0.608, GPT-6 Luna by 0.101
  • DeepSeek V4.1 Flash vs GPT-6 Sol
    0.507 vs 0.608, GPT-6 Sol by 0.101
  • DeepSeek V4.1 Flash vs GPT-5.6 Sol
    0.507 vs 0.605, GPT-5.6 Sol by 0.098
  • DeepSeek V4.1 Flash vs Muse Wonder
    0.507 vs 0.605, Muse Wonder by 0.098
  • DeepSeek V4.1 Flash vs Claude Opus 5
    0.507 vs 0.598, Claude Opus 5 by 0.091
  • DeepSeek V4.1 Flash vs Muse Spark 1.1
    0.507 vs 0.593, Muse Spark 1.1 by 0.086
  • DeepSeek V4.1 Flash vs Claude Sonnet 5
    0.507 vs 0.578, Claude Sonnet 5 by 0.071
  • DeepSeek V4.1 Flash vs GPT-5.6 Terra
    0.507 vs 0.577, GPT-5.6 Terra by 0.070
  • DeepSeek V4.1 Flash vs Claude Opus 4.8 (Opus 4.8 grader)
    0.507 vs 0.574, Claude Opus 4.8 (Opus 4.8 grader) by 0.067
  • DeepSeek V4.1 Flash vs Grok 4.7
    0.507 vs 0.567, Grok 4.7 by 0.060
  • DeepSeek V4.1 Flash vs Claude Opus 4.8
    0.507 vs 0.558, Claude Opus 4.8 by 0.051
  • DeepSeek V4.1 Flash vs GPT-5.6 Luna
    0.507 vs 0.557, GPT-5.6 Luna by 0.050
  • DeepSeek V4.1 Flash vs Gemini 4 Argon
    0.507 vs 0.551, Gemini 4 Argon by 0.044
  • DeepSeek V4.1 Flash vs Muse Spark
    0.507 vs 0.541, Muse Spark by 0.034
  • DeepSeek V4.1 Flash vs GPT-5.6 Sol (August)
    0.507 vs 0.540, GPT-5.6 Sol (August) by 0.033
  • DeepSeek V4.1 Flash vs Kimi K3
    0.507 vs 0.532, Kimi K3 by 0.025
  • DeepSeek V4.1 Flash vs GLM 5.3
    0.507 vs 0.528, GLM 5.3 by 0.021
  • DeepSeek V4.1 Flash vs Claude Opus 4.7
    0.507 vs 0.519, Claude Opus 4.7 by 0.012
  • DeepSeek V4.1 Flash vs GPT-5.5
    0.507 vs 0.518, GPT-5.5 by 0.011
  • DeepSeek V4.1 Flash vs Gemini 3.8 Flash
    0.507 vs 0.517, Gemini 3.8 Flash by 0.010
  • DeepSeek V4.1 Flash vs Grok 4.6
    0.507 vs 0.485, DeepSeek V4.1 Flash by 0.022
  • DeepSeek V4.1 Flash vs GPT-5.4
    0.507 vs 0.481, DeepSeek V4.1 Flash by 0.026
  • DeepSeek V4.1 Flash vs GPT-5
    0.507 vs 0.462, DeepSeek V4.1 Flash by 0.045
  • DeepSeek V4.1 Flash vs GPT-5.2
    0.507 vs 0.459, DeepSeek V4.1 Flash by 0.048
  • DeepSeek V4.1 Flash vs Claude Sonnet 4.6
    0.507 vs 0.442, DeepSeek V4.1 Flash by 0.065
  • DeepSeek V4.1 Flash vs GPT-5.6 Luna (August)
    0.507 vs 0.441, DeepSeek V4.1 Flash by 0.066
  • DeepSeek V4.1 Flash vs GPT-5.1
    0.507 vs 0.396, DeepSeek V4.1 Flash by 0.111
  • DeepSeek V4.1 Flash vs GPT-5.5 Instant
    0.507 vs 0.384, DeepSeek V4.1 Flash by 0.123
  • DeepSeek V4.1 Flash vs MAI-Thinking-1
    0.507 vs 0.350, DeepSeek V4.1 Flash by 0.157
  • DeepSeek V4.1 Flash vs Claude Haiku 4.5
    0.507 vs 0.322, DeepSeek V4.1 Flash by 0.185

Where the scores come from and how they are read is on the methodology page. The full ranking is on the leaderboard.