Best by task

Best for Reasoning: Claude Mythos Preview71.1 indexHighest reasoning index →
Best for Coding: Gemini 3.1 Pro21 arenaTop coding arena score →
Fastest LLM: Step-3.5-Flash629 c/sHighest P95 output throughput →
Cheapest Frontier: Nemotron 3 Nano (30B A3B)$0.06/1M inLowest price at frontier quality →
Largest Context: Grok 4 Fast2M tokensBiggest context window →
Best Open-Weight: GLM-5.291.2% GPQATop open-weight model →

LLM Leaderboard 2026

Name: LLM Leaderboard 2026 — AI Models with Live Pricing & Benchmarks
Creator: LLM Stats
License: https://llm-stats.com/legal/terms-of-service

Compare 300+ AI models by the LLM Stats Score — intelligence, speed and price, updated continuously from public benchmarks and live API metrics.

Select 2-4 models to compare

← Scroll →


Gemini 3.1 Pro	🇺🇸	Closed	1.0M	$2.50	$15.00	59c/s	2,139	55.4	51.9	41.3	35.6	–	38.7	30.6	14.2	30.5	25.5	27.8	94.3%	—	80.6%	77.1%	92.6%	—	85.9%	—	80.5%	—	69.2%	51.4%	—	—	—	—	—	—	26.3%	59.0%	33.5%	54.2%	-	27.9s	Jan. 2025	Feb. 2026	Google
Claude Fable 5	🇺🇸	Closed	1M	$10.00	$50.00	42c/s	2,138	67.0	52.3	58.9	–	–	50.2	34.5	–	49.4	51.7	32.5	—	—	95.0%	—	—	—	—	—	—	—	—	64.5%	—	—	—	—	—	—	—	—	—	80.0%	-	95.8s	-	Jun. 2026	Anthropic
Claude Opus 4.6	🇺🇸	Closed	1M	$5.00	$25.00	35c/s	2,121	57.0	49.3	42.9	34.3	44.9	33.2	32.4	33.9	44.4	41.9	13.0	91.3%	99.8%	80.8%	68.8%	91.1%	—	84.0%	77.4%	77.3%	—	62.7%	53.1%	—	72.7%	—	—	—	—	76.0%	—	—	—	-	4.9s	-	Feb. 2026	Anthropic
GPT-5.5	🇺🇸	Closed	1.1M	$5.00	$30.00	47c/s	2,088	58.7	48.9	49.3	32.7	31.5	43.5	37.9	26.5	39.7	32.9	–	93.6%	—	—	85.0%	—	—	84.4%	—	83.2%	—	75.3%	52.2%	—	—	55.6%	—	—	35.4%	74.0%	—	—	58.6%	-	13.3s	Dec. 2025	Apr. 2026	OpenAI
Claude Opus 4.7	🇺🇸	Closed	1M	$5.00	$25.00	32c/s	1,940	57.8	49.0	48.1	27.8	–	42.9	34.8	16.7	46.4	45.5	35.8	94.2%	—	87.6%	—	91.5%	—	79.3%	91.0%	—	—	77.3%	54.7%	—	—	—	—	—	—	—	—	—	64.3%	-	3.2s	-	Apr. 2026	Anthropic
GLM-5.1	🇨🇳	Open	200k	$1.40	$4.40	223c/s	1,758	50.2	42.1	41.1	26.9	–	36.3	26.9	–	36.0	34.1	–	86.2%	—	—	—	—	—	79.3%	—	—	—	71.8%	52.3%	—	—	40.7%	—	—	—	—	—	—	58.4%	754	712ms	-	Apr. 2026	ZAI
GPT-5.4	🇺🇸	Closed	1M	$2.50	$15.00	150c/s	1,742	54.1	47.8	41.8	28.4	36.0	35.7	34.6	19.1	40.1	36.6	37.8	92.8%	—	—	73.3%	—	—	82.7%	—	81.2%	—	67.2%	39.8%	—	—	54.6%	—	—	47.6%	—	—	—	57.7%	-	1.9s	-	Mar. 2026	OpenAI
Gemini 3 Flash	🇺🇸	Closed	1M	$0.50	$3.00	65c/s	1,720	47.5	46.3	29.4	–	–	30.9	22.9	9.6	30.3	18.8	42.0	90.4%	99.7%	78.0%	33.6%	91.8%	—	—	80.3%	81.2%	69.1%	57.4%	43.5%	68.7%	—	49.4%	—	—	—	22.1%	—	—	—	-	3.5s	Jan. 2025	Dec. 2025	Google
Gemini 3.5 Flash	🇺🇸	Closed	1.0M	$1.50	$9.00	393c/s	1,696	55.1	40.3	43.5	–	–	38.9	41.0	18.9	43.9	36.1	–	—	—	—	72.1%	—	—	—	84.2%	83.6%	—	83.6%	40.2%	—	—	56.5%	—	—	—	26.6%	—	—	55.1%	-	14.9s	Jan. 2026	May 2026	Google
Qwen3.7 Max	🇨🇳	Closed	1M	$1.25	$3.75	72c/s	1,692	57.6	51.3	45.6	–	–	33.6	34.8	–	48.6	50.6	60.5	92.4%	—	80.4%	—	90.3%	—	—	—	—	—	76.4%	41.4%	—	—	—	—	—	—	—	53.5%	—	60.6%	-	4.0s	-	May 2026	Qwen
Claude Sonnet 4.6	🇺🇸	Closed	200k	$3.00	$15.00	47c/s	1,681	48.1	42.9	35.0	22.5	36.1	31.5	27.7	13.6	43.7	40.7	24.6	89.9%	—	79.6%	58.3%	89.3%	—	74.7%	—	75.6%	—	61.3%	49.0%	—	72.5%	—	—	—	—	—	—	—	—	-	2.3s	-	Feb. 2026	Anthropic
Claude Opus 4.5	🇺🇸	Closed	-	-	-	102c/s	1,614	52.0	42.2	38.0	–	36.3	27.9	28.6	–	–	–	28.2	87.0%	—	80.9%	37.6%	90.8%	—	—	—	—	—	62.3%	—	—	66.3%	—	—	—	—	—	—	—	—	-	1.5s	Mar. 2025	Nov. 2025	Anthropic
Claude Opus 4.8	🇺🇸	Closed	1M	$5.00	$25.00	47c/s	1,598	63.4	50.7	52.3	36.5	–	47.8	40.8	31.8	44.2	46.5	26.4	93.6%	—	88.6%	—	—	—	84.3%	89.9%	—	87.9%	82.2%	57.9%	—	—	59.9%	—	—	—	—	—	—	69.2%	-	2.9s	-	May 2026	Anthropic
GLM-5	🇨🇳	Open	200k	$1.00	$3.20	–	1,586	49.3	–	34.5	24.3	–	–	24.3	–	–	–	–	—	—	77.8%	—	—	—	75.9%	—	—	—	67.8%	—	—	—	—	—	—	—	—	—	—	—	744	–	-	Feb. 2026	ZAI
Gemini 3 Pro	🇺🇸	Closed	-	-	-	–	1,579	48.4	46.9	31.1	–	–	31.9	18.4	13.1	–	–	50.0	91.9%	100.0%	76.2%	31.1%	91.8%	—	—	81.4%	81.0%	72.7%	—	45.8%	72.1%	—	—	—	—	—	26.3%	—	—	—	-	–	Jan. 2025	Nov. 2025	Google
Claude Sonnet 5	🇺🇸	Closed	1M	$3.00	$15.00	85c/s	1,568	62.1	52.4	49.5	33.4	–	44.6	37.6	–	44.2	45.1	28.8	—	—	85.2%	—	—	—	84.7%	88.3%	—	—	—	57.4%	—	—	54.3%	—	—	—	—	—	—	63.2%	-	211.2s	-	Jun. 2026	Anthropic
Kimi K2.6	🇨🇳	Open	262.1k	$0.75	$3.50	122c/s	1,557	54.9	45.5	41.4	36.3	–	36.0	29.8	–	35.9	33.0	–	90.5%	—	80.2%	—	—	—	86.3%	86.7%	80.1%	—	—	36.4%	—	—	50.0%	—	—	—	—	52.2%	27.9%	58.6%	1000	7.6s	-	Apr. 2026	MoonshotAI
GPT-5.2	🇺🇸	Closed	400k	$1.75	$14.00	47c/s	1,545	51.6	47.7	33.3	25.2	33.8	33.5	26.9	–	–	–	44.8	92.4%	100.0%	80.0%	52.9%	89.6%	—	65.8%	82.1%	79.5%	86.3%	60.6%	34.5%	—	—	46.3%	—	—	40.3%	—	—	—	—	-	50.1s	Aug. 2025	Dec. 2025	OpenAI
GLM-5.2	🇨🇳	Open	1.0M	$0.95	$3.00	99c/s	1,508	59.1	50.4	48.8	–	–	38.2	33.2	–	–	–	–	91.2%	—	—	—	—	—	—	—	—	—	76.8%	54.7%	—	—	48.2%	—	—	—	—	—	—	62.1%	753	14.6s	-	Jun. 2026	ZAI
Kimi K2.5	🇨🇳	Open	-	-	-	47c/s	1,462	47.9	44.0	30.5	28.5	–	32.6	13.1	34.9	45.7	46.0	46.5	87.6%	96.1%	76.8%	—	—	—	74.9%	77.5%	78.5%	—	—	50.2%	—	—	—	—	—	—	—	48.7%	—	50.7%	1000	2.9s	-	Jan. 2026	MoonshotAI
Claude Sonnet 4.5	🇺🇸	Closed	200k	$3.00	$15.00	55c/s	1,445	39.9	32.7	30.6	–	33.6	17.5	33.6	21.2	–	–	16.6	83.4%	87.0%	—	—	89.1%	—	—	—	—	—	—	—	—	61.4%	—	50.0%	86.2%	—	—	—	—	—	-	2.5s	Jan. 2025	Sep. 2025	Anthropic
GPT-5.4 mini	🇺🇸	Closed	400k	$0.75	$4.50	83c/s	1,442	41.1	33.5	31.4	–	23.6	26.2	23.1	7.3	36.0	26.4	33.4	88.0%	—	—	—	—	—	—	—	76.6%	—	57.7%	28.2%	—	—	42.9%	—	—	—	33.6%	—	—	54.4%	-	2.1s	Aug. 2025	Mar. 2026	OpenAI
GPT-5.3 Codex	🇺🇸	Closed	400k	$1.75	$14.00	151c/s	1,388	49.6	30.8	40.8	–	–	24.5	36.9	–	–	–	–	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	56.8%	-	2.8s	-	Feb. 2026	OpenAI
GPT-5.5 Instant	🇺🇸	Closed	400k	$5.00	$30.00	40c/s	1,375	41.0	20.5	–	–	–	27.3	–	–	–	–	30.4	85.6%	81.2%	—	—	—	—	—	81.6%	76.0%	—	—	—	—	—	—	—	—	—	—	—	—	—	-	1.7s	Aug. 2025	May 2026	OpenAI
DeepSeek-V4-Pro-Max	🇨🇳	Open	1.0M	$1.60	$3.20	80c/s	1,318	54.7	49.2	41.1	29.6	–	32.0	32.1	19.3	44.3	44.3	47.9	90.1%	—	80.6%	—	—	—	83.4%	—	—	—	73.6%	48.2%	57.9%	—	51.8%	—	—	—	—	—	—	55.4%	1600	18.6s	-	Apr. 2026	DeepSeek
GPT-5 High	🇺🇸	Closed	-	-	-	–	1,301	48.2	40.3	–	–	–	–	–	–	–	–	–	87.3%	94.6%	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	-	–	Sep. 2024	Aug. 2025	OpenAI
Qwen3.5-397B-A17B	🇨🇳	Open	-	-	-	–	1,296	47.7	44.3	28.9	24.6	23.4	25.9	22.4	37.8	42.2	42.6	51.1	88.4%	—	76.4%	—	88.5%	—	69.0%	—	—	—	—	28.7%	—	—	38.3%	—	—	—	—	—	—	—	397	–	-	Feb. 2026	Qwen
Gemma 4 31B	🇺🇸	Open	262.1k	$0.13	$0.38	160c/s	1,287	41.2	37.9	–	–	–	25.5	19.3	23.4	31.4	31.5	36.7	84.3%	—	—	—	88.4%	—	—	—	76.9%	—	—	26.5%	—	—	—	—	—	—	—	—	—	—	30.7	736ms	Jan. 2025	Apr. 2026	Google
GPT-5.1	🇺🇸	Closed	400k	$1.25	$10.00	166c/s	1,258	48.0	41.3	30.3	22.2	28.8	33.7	24.4	–	–	–	47.5	88.1%	94.0%	76.3%	—	—	85.4%	—	—	—	—	—	—	—	—	—	—	—	26.7%	—	—	—	—	-	1.5s	Sep. 2024	Nov. 2025	OpenAI
GPT-5.2 Codex	🇺🇸	Closed	400k	$1.75	$14.00	266c/s	1,252	47.3	33.9	36.0	–	–	–	24.5	–	–	–	–	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	56.4%	-	13.8s	-	Jan. 2026	OpenAI

Showing 1–30 of 324 models

LLM Leaderboard highlights

The LLM Leaderboard ranks 300+ AI models by intelligence, output speed, latency and per-token pricing, aggregated into the LLM Stats Score. Updated continuously from provider APIs and verified benchmarks. See the LLM Stats Score methodology for how rankings are computed.

Best for coding (Arena): Gemini 3.1 Pro (21.4 arena score)
Best on GPQA Diamond: Claude Mythos Preview (94.6%)
Best on AIME 2025: Gemini 3 Pro (100.0%)
Best on SWE-Bench Verified: Claude Fable 5 (95.0%)
Highest throughput: Step-3.5-Flash (629 tok/s)
Lowest latency: Qwen3 30B A3B (443.00 s)
Cheapest input: Nemotron 3 Nano (30B A3B) ($0.06 / 1M tok)
Largest context window: Grok 4 Fast (2.0M tokens)

FAQ

Common questions about the llm leaderboard

What is the best LLM right now?

Based on coding-arena performance — the most discriminating signal at the frontier — Gemini 3.1 Pro currently leads. For knowledge-heavy reasoning (GPQA Diamond), Claude Mythos Preview scores highest. Choose by axis rather than a single ranking — see the highlights above for per-metric leaders.

How does the LLM Leaderboard rank models?

Models are sorted by coding-arena score (when available), then by GPQA Diamond. Each row aggregates verified benchmark results, provider-reported pricing, and live performance metrics (output throughput and time-to-first-token) sampled across the major API providers. See the LLM Stats Score methodology for the full weighting and refresh cadence.

How many models are tracked?

This leaderboard tracks 324 canonical models across every major lab and provider. New releases typically appear within hours.

Where does pricing data come from?

Per-model input/output pricing is pulled from each provider's public API price list and verified against billing samples from the LLM Stats proxy. When a model is hosted by multiple providers, the cheapest available rate is shown by default.

How is performance measured?

Output throughput (tokens/second) and time-to-first-token are measured by routing standardized prompts through each provider's API and averaging over a 7-day rolling window. Numbers update hourly. Per-model splits live on each model detail page.

How often does the data update?

Pricing and model metadata revalidate every hour. Live performance metrics update on a 7-day rolling average. Benchmark scores update when a new verified result is published or a new evaluation lands on LLM Stats.