Vals MedCode (MedCode)

We show this table for reference; we do not rank on it.

Data verified October 29, 2026

Vals AI healthcare benchmark for whether models can support the medical billing process.

Evaluation results published by Vals AI. Published leaderboard aggregates and evaluation settings only. Dataset contents are not included.

MedCode score on MedCode — October 29, 2026

We mirror the published medcode score view for MedCode. Claude Opus 5 leads the public snapshot at 63.57%, followed by Gemini 3.1 Pro Preview (59.06%) and Gemini 4 Argon (58.80%). We do not use these results to rank models overall.

1Closed

Claude Opus 5

Anthropic

63.57%

Overall 76.05Context 1M

Gemini 3.1 Pro Preview

Google

high reasoning

59.06%

3Closed

Gemini 4 Argon

Google

high reasoning

58.80%

Overall 79.03Context 1M

107 modelsKnowledgeCurrentDisplay onlyUpdated October 29, 2026

MedCode score table (107 models)

MedCode score results for MedCode
Rank	Model / configuration	Score	Parameters (B)	Open / closed
1	Claude Opus 5Anthropic	63.57%	Not reported	Closed
2	Gemini 3.1 Pro PreviewGooglehigh reasoning	59.06%	Not reported	Unknown
3	Gemini 4 ArgonGooglehigh reasoning	58.80%	Not reported	Closed
4	Claude Fable 5Anthropic	56.07%	Not reported	Closed
5	Gemini 3 Flash PreviewGooglehigh reasoning	55.92%	Not reported	Unknown
6	Gemini 3.5 FlashGooglehigh reasoning	55.83%	Not reported	Closed
7	Claude Opus 4.7Anthropic	54.86%	Not reported	Closed
8	Claude Fable 5.1Anthropic	53.51%	Not reported	Closed
9	Gemini 3.7 FlashGooglehigh reasoning	53.39%	Not reported	Closed
10	Claude Opus 4.8Anthropic	53.22%	Not reported	Closed
11	Gemini 3.6 FlashGooglehigh reasoning	53.15%	Not reported	Closed
12	Claude Sonnet 5.5Anthropic	52.92%	Not reported	Closed
13	GPT-5.1OpenAIhigh reasoning	52.73%	Not reported	Closed
14	Gemini 3 Pro PreviewGooglehigh reasoning	52.20%	Not reported	Unknown
15	Claude Haiku 5.5Anthropic	51.54%	Not reported	Closed
16	Muse SparkMeta	51.31%	Not reported	Closed
17	Gemini 2.5 ProGoogle	50.59%	Not reported	Closed
18	Muse Wonder FullMetamax reasoning	49.93%	Not reported	Closed
19	Claude Opus 5.5Anthropic	49.80%	Not reported	Closed
20	GPT-5.2OpenAIxhigh reasoning	49.75%	Not reported	Closed
21	GPT-5OpenAIhigh reasoning	49.63%	Not reported	Unknown
22	Grok 4.7xAIxhigh reasoning	49.55%	Not reported	Closed
23	Kimi K3Moonshot AImax reasoning	49.36%	Not reported	Open
24	Muse Spark 1.2Metaxhigh reasoning	49.35%	Not reported	Closed
25	Claude Opus 4.5 ThinkingAnthropic	49.16%	Not reported	Closed
26	Claude Opus 4.6 (Adaptive)Anthropic	49.13%	Not reported	Closed
27	GPT-5.5OpenAIxhigh reasoning	49.10%	Not reported	Closed
28	GPT-6.1 SolOpenAImax reasoning	48.84%	Not reported	Closed
29	GPT-6 AstraOpenAImax reasoning	48.49%	Not reported	Closed
30	Claude Opus 4.6Anthropic	48.24%	Not reported	Closed
31	Gemini 3.8 FlashGooglehigh reasoning	48.13%	Not reported	Closed
32	Gemini 3.1 Flash Lite PreviewGooglehigh reasoning	47.60%	Not reported	Unknown
33	Claude Sonnet 5Anthropic	47.54%	Not reported	Closed
34	o3OpenAIhigh reasoning	47.29%	Not reported	Closed
35	Claude Opus 4.1 20250805 ThinkingAnthropic	47.23%	Not reported	Unknown
36	GPT-6 SolOpenAImax reasoning	47.07%	Not reported	Closed
37	MiniMax M3MiniMax	46.29%	Not reported	Open
38	Claude Opus 4.5Anthropic	45.17%	Not reported	Closed
39	MiMo-V2.6-ProXiaomi	44.97%	Not reported	Open
40	Grok 4.6xAIhigh reasoning	44.71%	Not reported	Closed
41	GPT-6 LunaOpenAImax reasoning	44.69%	Not reported	Closed
42	Claude Sonnet 4.5 ThinkingAnthropic	44.13%	Not reported	Closed
43	GPT-5.6 SolOpenAImax reasoning	43.97%	Not reported	Closed
44	Gemini 3.5 Flash-LiteGooglehigh reasoning	43.49%	Not reported	Closed
45	GPT-5.6 TerraOpenAIxhigh reasoning	43.41%	Not reported	Closed
46	Grok 4.5xAIhigh reasoning	43.29%	Not reported	Closed
47	Hy4 previewTencent	43.25%	Not reported	Open
48	GPT-5 miniOpenAIhigh reasoning	43.05%	Not reported	Closed
49	GLM-5.3Z.AImax reasoning	42.86%	Not reported	Open
50	DeepSeek V4 Pro 0813DeepSeekmax reasoning	42.47%	Not reported	Open
51	GPT-5.6 LunaOpenAImax reasoning	42.39%	Not reported	Closed
52	GLM-5.1Z.AI	41.60%	Not reported	Open
53	DeepSeek V4 Flash 0731DeepSeekhigh reasoning	41.41%	Not reported	Open
54	Claude Opus 4.1Anthropic	41.37%	Not reported	Unknown
55	GPT-5.4OpenAIxhigh reasoning	41.29%	Not reported	Closed
56	InklingThinking Machines Lab0.99 reasoning	41.19%	Not reported	Open
57	DeepSeek V4.1 Flash 1017DeepSeekhigh reasoning	41.17%	Not reported	Open
58	MiMo-V2.6-FlashXiaomi	41.06%	Not reported	Open
59	GPT-5.4 nanoOpenAIhigh reasoning	41.03%	Not reported	Closed
60	GLM-5.2Z.AI	40.77%	Not reported	Open
61	Mistral Large 4Mistralhigh reasoning	40.69%	Not reported	Open
62	Qwen3.8 MaxAlibaba	40.67%	Not reported	Open
63	Claude Sonnet 4.5Anthropic	40.57%	Not reported	Closed
64	Gemini 2.5 Flash Preview 09 2025Google	40.54%	Not reported	Unknown
65	DeepSeek V4 Pro 0813DeepSeekmax reasoning	40.45%	Not reported	Open
66	Gemini 2.5 Flash ThinkingGoogle	40.36%	Not reported	Unknown
67	Gemini 2.5 Flash Preview 09 2025 ThinkingGoogle	40.33%	Not reported	Unknown
68	Kimi K2.6Moonshot AI	40.14%	Not reported	Open
69	Kimi K2.5 ThinkingMoonshot AI	39.32%	Not reported	Unknown
70	Qwen3.7 MaxAlibaba	38.75%	Not reported	Closed
71	Nemotron 3 Ultra 550b A55bNvidia	38.62%	Not reported	Unknown
72	Gemini 2.5 FlashGoogle	38.42%	Not reported	Closed
73	Grok 4 0709xAI	38.08%	Not reported	Unknown
74	Grok 4.3xAI	38.07%	Not reported	Closed
75	Inkling-SmallThinking Machines Lab0.99 reasoning	37.89%	Not reported	Open
76	Grok 4 Fast (Reasoning)xAI	37.38%	Not reported	Closed
77	Qwen3.6 PlusAlibaba	36.89%	Not reported	Closed
78	Llama4 Maverick Instruct BasicFireworks AI	36.51%	Not reported	Unknown
79	Claude Sonnet 4 20250514 ThinkingAnthropic	34.96%	Not reported	Unknown
80	MiniMax M2.7MiniMax	34.44%	Not reported	Open
81	Gemini 2.5 Flash Lite Preview 09 2025 ThinkingGoogle	34.19%	Not reported	Unknown
82	MiniMax M2.1MiniMax	34.08%	Not reported	Unknown
83	Claude Sonnet 4Anthropic	33.94%	Not reported	Unknown
84	o4-miniOpenAIhigh reasoning	33.79%	Not reported	Closed
85	Mistral Medium 3.5Mistral AIhigh reasoning	33.75%	Not reported	Unknown
86	Qwen3.5 FlashAlibaba	33.00%	Not reported	Closed
87	GLM-4.7Z.AI	32.77%	Not reported	Open
88	Claude Haiku 4.5 ThinkingAnthropic	32.68%	Not reported	Closed
89	MiMo-V2.5-ProXiaomi	32.48%	Not reported	Closed
90	Ling 3.0 Flash 2607Ant	32.27%	Not reported	Unknown
91	Grok 4.20 0309 ReasoningxAI	32.16%	Not reported	Unknown
92	MiMo-V2.5Xiaomi	31.89%	Not reported	Closed
93	Qwen3 VL PlusAlibaba	31.65%	Not reported	Unknown
94	Qwen3 MaxAlibaba	31.37%	Not reported	Closed
95	Mercury 2.5Inceptionhigh reasoning	31.33%	Not reported	Closed
96	GPT-5 nanoOpenAIhigh reasoning	30.44%	Not reported	Closed
97	Grok 4 Fast Non ReasoningxAI	30.04%	Not reported	Unknown
98	Ling 3.0 Flash Af Rc3Ant	29.30%	Not reported	Unknown
99	Qwen3.8-27BAlibabaxhigh reasoning	28.70%	Not reported	Open
100	Grok 4.1 Fast Non ReasoningxAI	28.35%	Not reported	Unknown
101	Grok 4.1 Fast (Reasoning)xAI	28.08%	Not reported	Closed
102	Gemini 2.5 Flash-LiteGoogle	27.11%	Not reported	Closed
103	Gemini 2.5 Flash Lite Preview 09 2025Google	27.08%	Not reported	Unknown
104	Meta Llama Llama 4 Scout 17B 16E InstructTogether AI	23.31%	Not reported	Unknown
105	Laguna M.1Poolside	23.11%	Not reported	Closed
106	Laguna XS.2Poolside	21.25%	Not reported	Open
107	Command A Plus 05 2026Cohere	19.72%	Not reported	Unknown

How MedCode is shown here

BenchLM mirrors the public Vals AI MedCode leaderboard captured from https://www.vals.ai/benchmarks/medcode and updated by Vals on October 29, 2026. The snapshot preserves overall scores, uncertainty, latency, cost-per-test metadata, and task-level scores where Vals publishes them.

MedCode is display only on BenchLM. Vals proprietary or Vals-hosted aggregate views are useful context, but BenchLM does not use them as weighted ranking inputs or as a replacement for benchmark-native source records.

Snapshot

107 Vals rows1 task viewsprivate datasetTasks: OverallDisplay only

MedCode on Vals AI Vals methodology Vals home

The published MedCode snapshot places Claude Opus 5 first at 63.57%. The third row is 4.77 points behind. The broader top-10 range is 10.35 points, so the table still separates the published systems.

107 models have been evaluated on MedCode. The benchmark falls in the Knowledge category. We keep external benchmark mirrors separate from the weighted global scoring system, so these results remain source-specific evidence. MedCode is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.

About MedCode

Year

2026

Tasks

Medical billing support tasks

Format

Accuracy score

Difficulty

Professional healthcare administration

BenchLM mirrors the public Vals MedCode leaderboard as display-only healthcare evidence.

MedCode Public benchmark source

Freshness and provenance

Version

MedCode 2026

Refresh cadence

Quarterly

Staleness state

Current

Question availability

Public benchmark set

CurrentDisplay only

BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.

Questions

What does MedCode measure?

Vals AI healthcare benchmark for whether models can support the medical billing process.

Which model leads the published MedCode snapshot?

Claude Opus 5 currently leads the published MedCode snapshot with 63.57% medcode score. BenchLM shows this benchmark for display only and does not use it in overall rankings.

How many models are evaluated on MedCode?

The October 29, 2026 snapshot contains 107 AI models.

Last updated: October 29, 2026 · mirrored from the public benchmark leaderboard

Know when it’s worth switching models

The model to choose, the cheaper alternative, and the release we would wait on.

Read a sample issue

Join 7,000+ subscribers.

One email each week. Unsubscribe anytime.