SciCode

AI benchmark results and model performance

Type: percentage

44.15%

24.70% - 60.10%

24.70% - Solar Pro 3 (2026-03-23)

Scores Over Time

Individual benchmark scores plotted by date.

Organisation	Model	Reported	Top Score	Info	Self Reported	Source
Upstage	Solar Pro 3 (2026-03-23)	23 Mar 2026	24.70%	Artificial Analysis structured model metrics	No	Source
Nvidia	Nemotron 3.5 Lightning 30B A3B	11 Aug 2026	32.60%	NVIDIA release evaluation suite	Yes	Source
Nvidia	Nemotron 3 Nano 30B A3B	15 Dec 2025	33.30%	-	Yes	Source
MiniMax	MiniMax M2	27 Oct 2025	36%	-	Yes	-
MiniMax	MiniMax M2 Her	24 Jan 2026	36%	inferred modality/version alias from minimax-m2	Yes	-
z.AI	GLM 4.5 Air	28 Jul 2025	37.30%	-	Yes	Source
Inception	Mercury 2	24 Feb 2026	38%	-	Yes	Source
MiniMax	MiniMax M2.1	23 Dec 2025	39%	-	Yes	Source
z.AI	GLM 4.5	28 Jul 2025	41.70%	-	Yes	Source
Nvidia	Nemotron 3 Super	11 Mar 2026	42.05%	-	Yes	Source
z.AI	GLM 5 Turbo	15 Mar 2026	43.60%	Artificial Analysis structured model metrics	No	Source
Meta	Muse Glimmer 30B	10 Aug 2026	43.60%	High reasoning	Yes	Source
Moonshot	Kimi K2 Thinking	06 Nov 2025	44.80%	inferred alias from kimi-k2-thinking-0905	Yes	Source
SpaceXAI	Grok 4.20	17 Feb 2026	45.60%	Artificial Analysis structured model metrics	No	Source
Moonshot	Kimi K2.5	27 Jan 2026	48.70%	-	Yes	Source
ByteDance	Seed 2.0 Pro	14 Feb 2026	52.10%	Seed2 official benchmark table \| Scicode	Yes	Source
ByteDance	Seed 2.0 Lite	14 Feb 2026	52.40%	Seed2 official benchmark table \| Scicode	Yes	Source
Qwen	Qwen 3.7 Max	21 May 2026	53.50%	-	Yes	Source
Sakana AI	Fugu Ultra	22 Jun 2026	58.70%	Sakana Fugu product page benchmark table.	Yes	Source
Google	Gemini 3.1 Pro Preview	19 Feb 2026	59%	-	Yes	-
Sakana AI	Fugu	22 Jun 2026	60.10%	Sakana Fugu product page benchmark table.	Yes	Source