SciCode

Coding

Source ↗ Higher is better

Rank	Model	Provider	Score
1	Claude Opus 5.5 (Adaptive Reasoning, Max Effort, Default Fallback)	Anthropic	66.9%
2	Claude Opus 5.5 (Adaptive Reasoning, Xhigh Effort, Default Fallback)	Anthropic	65.0%
3	Claude Fable 5.1 (Adaptive Reasoning, Max Effort, Default Fallback)	Anthropic	63.1%
4	Muse Wonder Full (Max)	Meta	62.6%
5	Gemini 4 Argon (High)	Google	61.8%
6	Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)	Anthropic	61.0%
7	Claude Sonnet 5.5 (Adaptive Reasoning, Max Effort, Default Fallback)	Anthropic	61.0%
8	Xiaomi: MiMo-V2.6-Pro	Xiaomi	60.9%
9	Claude Fable 5.1 (Adaptive Reasoning, Xhigh Effort, Default Fallback)	Anthropic	60.9%
10	Claude Opus 5.5 (Adaptive Reasoning, High Effort, Default Fallback)	Anthropic	60.4%
11	Muse Wonder (Max)	Meta	60.3%
12	Gemini 3.7 Flash (medium)	Google	59.8%
13	Muse Spark 1.3 (xhigh)	Meta	59.7%
14	Kimi K3 (max)	Moonshot AI	59.5%
15	Claude Opus 5.5 (Adaptive Reasoning, Medium Effort, Default Fallback)	Anthropic	59.3%
16	GLM-5.3 (max)	Z.ai	59.0%
17	StepFun: Step 5 Preview	Stepfun	58.9%
18	Muse Spark 1.1 (xhigh)	Meta	58.8%
19	Muse Spark 1.3 (max)	Meta	58.8%
20	Claude Fable 5.1 (Adaptive Reasoning, High Effort, Default Fallback)	Anthropic	58.7%
21	Google: Gemini 3.1 Pro Preview	Google	58.7%
22	Claude Opus 5.5 (Adaptive Reasoning, Low Effort, Default Fallback)	Anthropic	58.6%
23	Grok 4.7 (high)	SpaceXAI	57.8%
24	GPT-5.6 Sol (high)	OpenAI	57.8%
25	GPT-6 Sol (max)	OpenAI	57.6%
26	Muse Spark 1.2 (xhigh)	Meta	57.4%
27	Grok 4.7 (xhigh)	SpaceXAI	57.4%
28	GPT-5.6 Sol (medium)	OpenAI	57.4%
29	Claude Sonnet 5.5 (Adaptive Reasoning, Xhigh Effort, Default Fallback)	Anthropic	57.3%
30	Gemini 3.7 Flash (high)	Google	57.2%
31	GPT-5.6 Sol (max)	OpenAI	57.1%
32	GPT-5.6 Sol (xhigh)	OpenAI	57.1%
33	Claude Fable 5.1 (Adaptive Reasoning, Low Effort, Default Fallback)	Anthropic	56.7%
34	Gemini 3.8 Flash (high)	Google	56.6%
35	Grok 4.6 (high)	SpaceXAI	56.5%
36	GPT-6 Astra (max)	OpenAI	56.5%
37	Claude Opus 5 (Adaptive Reasoning, Max Effort)	Anthropic	56.4%
38	GPT-5.6 Sol (low)	OpenAI	56.4%
39	Claude Fable 5.1 (Adaptive Reasoning, Medium Effort, Default Fallback)	Anthropic	56.4%
40	GPT-5.5 (high)	OpenAI	56.1%
41	Grok 4.6 (medium)	SpaceXAI	55.9%
42	GPT-6.1 Sol (high)	OpenAI	55.8%
43	GPT-5.5 (xhigh)	OpenAI	55.8%
44	Claude Opus 5 (Adaptive Reasoning, Xhigh Effort)	Anthropic	55.7%
45	Gemini 3.7 Flash (low)	Google	55.7%
46	GPT-6 Astra (xhigh)	OpenAI	55.7%
47	GPT-6.1 Sol (xhigh)	OpenAI	55.7%
48	Claude Opus 5 (Adaptive Reasoning, High Effort)	Anthropic	55.4%
49	GPT-6 Astra (high)	OpenAI	55.4%
50	Gemini 3.8 Flash (medium)	Google	55.1%

Access this data

JSON GET /api/v1

[
  {
    "rank": 1,
    "model": "Claude Opus 5.5 (Adaptive Reasoning, Max Effort, Default Fallback)",
    "provider": "Anthropic",
    "score": "66.9"
  },
  {
    "rank": 2,
    "model": "Claude Opus 5.5 (Adaptive Reasoning, Xhigh Effort, Default Fallback)",
    "provider": "Anthropic",
    "score": "65.0"
  },
  {
    "rank": 3,
    "model": "Claude Fable 5.1 (Adaptive Reasoning, Max Effort, Default Fallback)",
    "provider": "Anthropic",
    "score": "63.1"
  }
]

Curl

$ curl -s https://rightrank.ai/api/v1/rankings/scicode.json | jq '.[:3]'

CLI brew install rightrank

$ rightrank top scicode --limit 5
$ rightrank compare claude-opus-4.6 gpt-5.4-pro --category scicode
$ rightrank value scicode  # best performance per dollar

Claude Code Skill Skill

# In Claude Code, use the RightRank skill:
/rightrank best model for scicode
/rightrank compare top 5 by value
/rightrank cheapest model scoring above 80 on scicode