LLM leaderboard for 2026

Compare models using publicly reported benchmark results and find the right one for reasoning, coding, tool use, browsing, personal assistance, and everyday tasks.

Updated: Nov 5, 2026Data as of: Nov 5, 2026

Best Overall

Based on Humanity’s Last Exam, one of the hardest AI benchmarks.

70.0K
Kimi K3.1 ProRank 1: Kimi K3.1 Pro by Moonshot AI, 70 percent. With tools. Stable HLE — full set. Max effort; temperature=1.0; top_p=0.95; full 2500-question set.. Reported result source for Kimi K3.1 Pro
67.7A
Claude Opus 5.5Rank 2: Claude Opus 5.5 by Anthropic, 67.7 percent. With tools. Stable HLE — full set. Provider benchmark table; tools enabled. Cross-model settings may differ.. Reported result source for Claude Opus 5.5
65.6A
Claude Fable 5.1Rank 3: Claude Fable 5.1 by Anthropic, 65.6 percent. With tools. Stable HLE — full set. Provider benchmark table; tools enabled. Cross-model settings may differ.. Reported result source for Claude Fable 5.1
64.7A
Claude Mythos PreviewRank 4: Claude Mythos Preview by Anthropic, 64.7 percent. With tools. Stable HLE — full set. . Reported result source for Claude Mythos Preview
64.5A
Claude Mythos 5Rank 5: Claude Mythos 5 by Anthropic, 64.5 percent. With tools. Stable HLE — full set. Full 2,500 questions; thinking auto; total 1M-token cap; no context compaction; Opus 4.6 judge. Tools variant: search, fetch, programmatic tool calling and code execution.. Reported result source for Claude Mythos 5
64.5A
Claude Sonnet 5.5Rank 6: Claude Sonnet 5.5 by Anthropic, 64.5 percent. With tools. Stable HLE — full set. . Reported result source for Claude Sonnet 5.5
64.5O
GPT-5.6 SolRank 7: GPT-5.6 Sol by OpenAI, 64.5 percent. With tools. HLE — source does not specify subset. temperature=1.0; top_p=0.95; max generation 163840 tokens; max context 300000 tokens; context management; GPT-5.6 Luna medium judge.. Reported result source for GPT-5.6 Sol
63.9D
DeepSeek V4.1 FlashRank 8: DeepSeek V4.1 Flash by DeepSeek, 63.9 percent. With tools. HLE — source does not specify subset. . Reported result source for DeepSeek V4.1 Flash
63.8A
Claude Fable 5Rank 9: Claude Fable 5 by Anthropic, 63.8 percent. With tools. Stable HLE — full set. . Reported result source for Claude Fable 5
63.6A
Claude Opus 5Rank 10: Claude Opus 5 by Anthropic, 63.6 percent. With tools. Stable HLE — full set. Provider benchmark table; tools enabled. Cross-model settings may differ.. Reported result source for Claude Opus 5
62.5Z
GLM-5.3Rank 11: GLM-5.3 by Z.ai, 62.5 percent. With tools. HLE — source does not specify subset. temperature=1.0; top_p=0.95; max generation 163840 tokens; max context 300000 tokens; context management; GPT-5.6 Luna medium judge.. Reported result source for GLM-5.3
62.1M
Muse Spark 1.1Rank 12: Muse Spark 1.1 by Meta, 62.1 percent. With tools. Stable HLE — full set. Full 2500 questions; o3-mini judge; bash and browser tools; source describes domain blocklist.. Reported result source for Muse Spark 1.1
60.0D
DeepSeek V4 Pro 0813Rank 13: DeepSeek V4 Pro 0813 by DeepSeek, 60 percent. With tools. HLE — source does not specify subset. . Reported result source for DeepSeek V4 Pro 0813
59.8K
Kimi K3Rank 14: Kimi K3 by Moonshot AI, 59.8 percent. With tools. HLE — source does not specify subset. temperature=1.0; top_p=0.95; max generation 163840 tokens; max context 300000 tokens; context management; GPT-5.6 Luna medium judge.. Reported result source for Kimi K3
58.8D
DeepSeek V4.1 Flash 1017Rank 15: DeepSeek V4.1 Flash 1017 by DeepSeek, 58.8 percent. With tools. Stable HLE — full set. Max effort; competitor result in Moonshot AI report.. Reported result source for DeepSeek V4.1 Flash 1017
58.7O
GPT-5.4 ProRank 16: GPT-5.4 Pro by OpenAI, 58.7 percent. With tools. Stable HLE — full set. Reasoning effort xhigh; research environment.. Reported result source for GPT-5.4 Pro
58.0M
Muse SparkRank 17: Muse Spark by Meta, 58 percent. Tools unspecified. HLE — source does not specify subset. Contemplating mode; parallel agents. Tool access not explicit in cited prose.. Reported result source for Muse Spark
57.9A
Claude Opus 4.8Rank 18: Claude Opus 4.8 by Anthropic, 57.9 percent. With tools. Stable HLE — full set. . Reported result source for Claude Opus 4.8
57.4A
Claude Haiku 5.5Rank 19: Claude Haiku 5.5 by Anthropic, 57.4 percent. With tools. Stable HLE — full set. Haiku 5.5 launch comparison; reported no-tools and with-tools configurations.. Reported result source for Claude Haiku 5.5
57.4A
Claude Sonnet 5Rank 20: Claude Sonnet 5 by Anthropic, 57.4 percent. With tools. Stable HLE — full set. Full 2,500 questions; thinking auto; total 1M-token cap; no context compaction; Opus 4.6 judge. Tools variant: search, fetch, programmatic tool calling and code execution.. Reported result source for Claude Sonnet 5

Agent Index vs Cost: Find the best value for knowledge work

Compare AI models on knowledge work capability and token cost, using our seven-benchmark Agent Index.

LLM heatmap

See which models lead across practical tasks and general benchmarks.

Business deliverables

Claude Sonnet 5.5

Professional document reasoning

GPT-6 Astra

Task objective completion

Gemini 4 Carbon

Knowledge work

Claude Opus 5.5

Data analysis

Gemini 4 Argon

Document and presentation creation

Claude Opus 5.5

Long-running agentic work

Claude Fable 5.1

Research reports

Muse Wonder 1.0

Workflow completion

Gemini 4 Carbon

Financial analysis and due diligence

Claude Opus 5.5

Tool use

Muse Spark

Web research

GPT-6 Astra

Business deliverablesReported benchmark setup

Claude Sonnet 5.557 models · Higher is better

Best models per task

Best in Reasoning

GPQA Diamond

GPT-6 Astra by OpenAI96.3%
GPT-6 Astra (Xhigh); 198-question Diamond subset; independent AA evaluation. Source for GPT-6 Astra on GPQA Diamond
Gemini 3.8 Flash by Google95.3%
Gemini 3.8 Flash (High); 198-question Diamond subset; independent AA evaluation. Source for Gemini 3.8 Flash on GPQA Diamond
Gemini 3.1 Pro by Google94.1%
Gemini 3.1 Pro Preview; 198-question Diamond subset; independent AA evaluation. Source for Gemini 3.1 Pro on GPQA Diamond
GPT-5.6 Sol by OpenAI94.1%
GPT-5.6 Sol (Max); 198-question Diamond subset; independent AA evaluation. Source for GPT-5.6 Sol on GPQA Diamond
Muse Spark 1.3 by Meta94.1%
Muse Spark 1.3 (Xhigh); 198-question Diamond subset; independent AA evaluation. Source for Muse Spark 1.3 on GPQA Diamond

Best in Coding

DeepSWE v1.1

Claude Opus 5 by Anthropic74%
mini-swe-agent; effort=max; ±4 percentage points. Source for Claude Opus 5 on DeepSWE v1.1
Gemini 3.8 Flash by Google74%
mini-swe-agent; effort=high; ±1 percentage points. Source for Gemini 3.8 Flash on DeepSWE v1.1
GPT-6 Astra by OpenAI74%
mini-swe-agent; effort=xhigh; ±3 percentage points. Source for GPT-6 Astra on DeepSWE v1.1
GPT-5.6 Sol by OpenAI73%
mini-swe-agent; effort=max; ±3 percentage points. Source for GPT-5.6 Sol on DeepSWE v1.1
Claude Fable 5 by Anthropic70%
mini-swe-agent; effort=xhigh; ±3 percentage points. Source for Claude Fable 5 on DeepSWE v1.1

Best in Terminal Use

Terminal-Bench 4.0

Claude Opus 5.5 by Anthropic64.8%
Claude Code; effort=max; 64.8% ± 3.1%. Source for Claude Opus 5.5 on Terminal-Bench 4.0
Claude Sonnet 5.5 by Anthropic61.8%
Claude Code; effort=max; 61.8% ± 2.9%. Source for Claude Sonnet 5.5 on Terminal-Bench 4.0
GPT-6 Astra by OpenAI58.2%
Codex; effort=max; 58.2% ± 2.8%. Source for GPT-6 Astra on Terminal-Bench 4.0
GPT-6.1 Sol by OpenAI58.2%
Codex; effort=max; 58.2% ± 3.1%. Source for GPT-6.1 Sol on Terminal-Bench 4.0
Claude Fable 5.1 by Anthropic57.9%
Claude Code; effort=max; 57.9% ± 3.8%. Source for Claude Fable 5.1 on Terminal-Bench 4.0

Best in Web Research

BrowseComp

GPT-6 Astra by OpenAI91.5%
OpenAI GPT-6 Astra launch comparison; best reported reasoning setting; see source evaluation footnotes. Source for GPT-6 Astra on BrowseComp
Claude Opus 5 by Anthropic90.8%
OpenAI GPT-6 Astra launch comparison; best reported reasoning setting; see source evaluation footnotes. Source for Claude Opus 5 on BrowseComp
GPT-5.6 Sol by OpenAI90.4%
OpenAI GPT-6 Astra launch comparison; best reported reasoning setting; see source evaluation footnotes. Source for GPT-5.6 Sol on BrowseComp
Claude Fable 5 by Anthropic87.4%
OpenAI GPT-6 Astra launch comparison; best reported reasoning setting; see source evaluation footnotes. Source for Claude Fable 5 on BrowseComp

Best in Work Automations

AutomationBench 1.0.6

Gemini 4 Carbon by Google53.88%
Gemini 4 Carbon (High); private held-out 657 tasks; API mode; 50-step cap; strict completion. Source for Gemini 4 Carbon on AutomationBench 1.0.6
Gemini 4 Argon by Google51.29%
Gemini 4 Argon (High); private held-out 657 tasks; API mode; 50-step cap; strict completion. Source for Gemini 4 Argon on AutomationBench 1.0.6
Claude Sonnet 5.5 by Anthropic44.75%
Claude Sonnet 5.5 (default fallbacks, Max); private held-out 657 tasks; API mode; 50-step cap; strict completion. Source for Claude Sonnet 5.5 on AutomationBench 1.0.6
Claude Opus 5.5 by Anthropic42.47%
Claude Opus 5.5 (default fallbacks, Max); private held-out 657 tasks; API mode; 50-step cap; strict completion. Source for Claude Opus 5.5 on AutomationBench 1.0.6
GPT-6 Astra by OpenAI41.40%
GPT-6 Astra (Max); private held-out 657 tasks; API mode; 50-step cap; strict completion. Source for GPT-6 Astra on AutomationBench 1.0.6

Best in Knowledge Work

GDPval-AA v2.1

Claude Opus 5.5 by Anthropic1,866 Elo
Claude Opus 5.5 (Max, Default Fallback); 220 tasks; Stirrup agent; pairwise Elo. Source for Claude Opus 5.5 on GDPval-AA v2.1
Kimi K3.1 Pro by Moonshot AI1,858 Elo
Kimi K3.1 Pro; 220 tasks; Stirrup agent; pairwise Elo. Source for Kimi K3.1 Pro on GDPval-AA v2.1
Claude Sonnet 5.5 by Anthropic1,838 Elo
Claude Sonnet 5.5 (Max, Default Fallback); 220 tasks; Stirrup agent; pairwise Elo. Source for Claude Sonnet 5.5 on GDPval-AA v2.1
Claude Fable 5.1 by Anthropic1,763 Elo
Claude Fable 5.1 (Max, Default Fallback); 220 tasks; Stirrup agent; pairwise Elo. Source for Claude Fable 5.1 on GDPval-AA v2.1
DeepSeek V4.1 Flash 1017 by DeepSeek1,746 Elo
DeepSeek V4.1 Flash 1017 (Max); 220 tasks; Stirrup agent; pairwise Elo. Source for DeepSeek V4.1 Flash 1017 on GDPval-AA v2.1

Fastest and most affordable models

Best reported provider per model for each metric.

Fastest models

tokens/sec

Output speed · higher is better

DeepSeek V4.1 Flash 1017 by DeepSeekvia Inco (FAST)1,075.3DeepSeek V4.1 Flash 1017 (max) (FAST) · Artificial Analysis: DeepSeek V4.1 Flash 1017 (max) (FAST); host Inco (FAST); standard uncached input/output pricing. TTFT is the first streamed token, which may be reasoning. Retrieval time is not measurement time; region/concurrency/quantization not specified in this record.
DeepSeek V4.1 Flash by DeepSeekvia Baseten (FAST)830DeepSeek V4.1 Flash (max) (FAST) · Artificial Analysis: DeepSeek V4.1 Flash (max) (FAST); host Baseten (FAST); standard uncached input/output pricing. TTFT is the first streamed token, which may be reasoning. Retrieval time is not measurement time; region/concurrency/quantization not specified in this record.
MiniMax M3.1 Flash by MiniMaxvia Inco812.6MiniMax-M3.1-Flash · Artificial Analysis: MiniMax-M3.1-Flash; host Inco; standard uncached input/output pricing. TTFT is the first streamed token, which may be reasoning. Retrieval time is not measurement time; region/concurrency/quantization not specified in this record.
GLM-5.3 by Z.aivia Systalyze688.2GLM-5.3 (max) · Artificial Analysis: GLM-5.3 (max); host Systalyze; standard uncached input/output pricing. TTFT is the first streamed token, which may be reasoning. Retrieval time is not measurement time; region/concurrency/quantization not specified in this record.
Kimi K3 by Moonshot AIvia Inco (FAST)481.7Kimi K3 (max) (FAST) · Artificial Analysis: Kimi K3 (max) (FAST); host Inco (FAST); standard uncached input/output pricing. TTFT is the first streamed token, which may be reasoning. Retrieval time is not measurement time; region/concurrency/quantization not specified in this record.

Lowest latency

TTFT

Time to first token · lower is better

DeepSeek V4.1 Flash 1017 by DeepSeekvia Baseten (FAST)0.182sDeepSeek V4.1 Flash 1017 (max) (FAST) · Artificial Analysis: DeepSeek V4.1 Flash 1017 (max) (FAST); host Baseten (FAST); standard uncached input/output pricing. TTFT is the first streamed token, which may be reasoning. Retrieval time is not measurement time; region/concurrency/quantization not specified in this record.
DeepSeek V4.1 Flash by DeepSeekvia Together AI0.311sDeepSeek V4.1 Flash (max) · Artificial Analysis: DeepSeek V4.1 Flash (max); host Together AI; standard uncached input/output pricing. TTFT is the first streamed token, which may be reasoning. Retrieval time is not measurement time; region/concurrency/quantization not specified in this record.
MiniMax M3.1 Flash by MiniMaxvia Together AI0.584sMiniMax-M3.1-Flash · Artificial Analysis: MiniMax-M3.1-Flash; host Together AI; standard uncached input/output pricing. TTFT is the first streamed token, which may be reasoning. Retrieval time is not measurement time; region/concurrency/quantization not specified in this record.
GLM-5.3 by Z.aivia Together AI0.597sGLM-5.3 (max) · Artificial Analysis: GLM-5.3 (max); host Together AI; standard uncached input/output pricing. TTFT is the first streamed token, which may be reasoning. Retrieval time is not measurement time; region/concurrency/quantization not specified in this record.
MiniMax M3 by MiniMaxvia MiniMax0.699sMiniMax-M3 · Explicit first-party AA measurement; reasoning configuration retained. Cross-provider medians are excluded. Retrieval time is not measurement time.

Most affordable

$/1M tokens

50% input + 50% output · lower is better

DeepSeek V4.1 Flash 1017 by DeepSeekvia DeepInfra$0.196Input $0.098/1M · Output $0.294/1M · Artificial Analysis: DeepSeek V4.1 Flash 1017 (max); host DeepInfra; standard uncached input/output pricing. TTFT is the first streamed token, which may be reasoning. Retrieval time is not measurement time; region/concurrency/quantization not specified in this record.
DeepSeek V4.1 Flash by DeepSeekvia DeepInfra$0.28Input $0.14/1M · Output $0.42/1M · Artificial Analysis: DeepSeek V4.1 Flash (max); host DeepInfra; standard uncached input/output pricing. TTFT is the first streamed token, which may be reasoning. Retrieval time is not measurement time; region/concurrency/quantization not specified in this record.
Claude Haiku 5.5 by Anthropicvia Anthropic$0.30Input $0.1/1M · Output $0.5/1M · USD per 1M tokens; uncached standard input and generated output. Configuration records context price tier. No speed or TTFT inferred.
GPT-6 Luna by OpenAIvia OpenAI$0.30Input $0.1/1M · Output $0.5/1M · USD per 1M tokens; uncached standard input and generated output. Configuration records context price tier. No speed or TTFT inferred.
GLM-5.3-Flash by Z.aivia Z.ai$0.325Input $0.15/1M · Output $0.5/1M · Explicit first-party AA measurement; reasoning configuration retained. Cross-provider medians are excluded. Retrieval time is not measurement time.

Model comparison

Compare each model and provider offering. Costs are per 1M tokens.

Download CSV

Model specifications and provider-specific token pricing, time to first token, and generation speed. A dash means no published value is recorded.

Claude Fable 5	Anthropic	1,000,000	$10.00	$50.00	128,000	117.332	68.168
Claude Fable 5	Azure	1,000,000	$10.00	$50.00	128,000	115.288	68.983
Claude Fable 5	Google	1,000,000	$10.00	$50.00	128,000	106.118	70.091
Claude Fable 5.1	Anthropic	1,000,000	$10.00	$50.00	128,000	256.491	69.417
Claude Haiku 5.5	Anthropic	—	$0.50	$2.50	—	—	—
Claude Haiku 5.5	Anthropic	—	$0.10	$0.50	—	—	—
Claude Mythos 5	Anthropic	—	$10.00	$50.00	128,000	—	—
Claude Mythos Preview	Anthropic	—	$25.00	$125.00	—	—	—
Claude Opus 4.6	Amazon	1,000,000	$5.00	$25.00	128,000	1.35	58.564
Claude Opus 4.6	Anthropic	1,000,000	$5.00	$25.00	128,000	1.778	39.935
Claude Opus 4.6	Azure	1,000,000	$5.00	$25.00	128,000	2.086	40.802
Claude Opus 4.6	Google	1,000,000	$5.00	$25.00	128,000	1.898	42.424
Claude Opus 4.7	Amazon	1,000,000	$5.00	$25.00	128,000	14.738	68.472
Claude Opus 4.7	Anthropic	1,000,000	$5.00	$25.00	128,000	33.055	53.476
Claude Opus 4.7	Azure	1,000,000	$5.00	$25.00	128,000	33.588	53.379
Claude Opus 4.7	Google	1,000,000	$5.00	$25.00	128,000	23.76	68.4
Claude Opus 4.8	Amazon	1,000,000	$5.00	$25.00	128,000	49.238	70.462
Claude Opus 4.8	Anthropic	1,000,000	$5.00	$25.00	128,000	56.523	65.881
Claude Opus 4.8	Google	1,000,000	$5.00	$25.00	128,000	57.41	67.892
Claude Opus 5	Amazon	1,000,000	$5.00	$25.00	128,000	88.226	56.473
Claude Opus 5	Anthropic	1,000,000	$5.00	$25.00	128,000	73.708	54.128
Claude Opus 5	Google	1,000,000	$5.00	$25.00	128,000	60.427	57.082
Claude Opus 5.5	Amazon	1,000,000	$4.00	$20.00	128,000	740.928	97.19
Claude Opus 5.5	Anthropic	1,000,000	$4.00	$20.00	128,000	640.87	95.521
Claude Opus 5.5	Azure	1,000,000	$4.00	$20.00	128,000	695.618	96.187
Claude Opus 5.5	Google	1,000,000	$4.00	$20.00	128,000	711.139	95.728
Claude Sonnet 5	Amazon	1,000,000	$2.00	$10.00	128,000	225.112	82.708
Claude Sonnet 5	Anthropic	1,000,000	$2.00	$10.00	128,000	167.136	82.915
Claude Sonnet 5	Azure	1,000,000	$2.00	$10.00	128,000	231.315	83.274
Claude Sonnet 5	Google	1,000,000	$2.00	$10.00	128,000	210.089	79.391
Claude Sonnet 5.5	Amazon	1,000,000	$2.00	$10.00	128,000	529.016	141.843
Claude Sonnet 5.5	Anthropic	1,000,000	$2.00	$10.00	128,000	514.907	139.308
Claude Sonnet 5.5	Azure	1,000,000	$2.00	$10.00	128,000	497.061	134.954
Claude Sonnet 5.5	Google	1,000,000	$2.00	$10.00	128,000	522.437	144.019
DeepSeek V4 Pro 0813	Baseten	1,000,000	$1.32	$3.96	393,216	1.589	143.882
DeepSeek V4 Pro 0813	Databricks	1,000,000	$1.32	$3.96	393,216	—	—
DeepSeek V4 Pro 0813	DeepInfra	1,000,000	$1.30	$2.60	393,216	1.129	78.412
DeepSeek V4 Pro 0813	DeepSeek	1,048,576	$1.32	$3.96	393,216	1.793	89.952
DeepSeek V4 Pro 0813	DigitalOcean	1,048,576	$1.32	$3.96	393,216	0.832	57.677
DeepSeek V4 Pro 0813	GMI	1,048,576	$1.056	$3.168	393,216	3.152	71.407
DeepSeek V4 Pro 0813	Nebius	1,048,576	$1.32	$3.96	393,216	1.602	172.165
DeepSeek V4 Pro 0813	Novita	1,048,576	$1.32	$3.96	393,216	2.123	72.307
DeepSeek V4 Pro 0813	SiliconFlow	1,048,576	$1.32	$3.96	393,216	2.088	64.823
DeepSeek V4 Pro 0813	Systalyze	1,048,576	$0.50	$2.00	393,216	1.894	305.002
DeepSeek V4.1 Flash	Baseten	1,000,000	$0.30	$1.20	—	0.63	244.981
DeepSeek V4.1 Flash	Baseten (FAST)	1,048,576	$0.60	$2.40	—	0.94	830.034
DeepSeek V4.1 Flash	Bitdeer AI	1,000,000	$0.144	$0.576	—	1.515	289.729
DeepSeek V4.1 Flash	CoreWeave	1,000,000	$0.20	$0.65	—	0.997	297.861
DeepSeek V4.1 Flash	Crusoe	1,048,576	$0.29	$1.20	—	0.771	205.633
DeepSeek V4.1 Flash	Databricks	1,000,000	$0.30	$1.20	—	0.91	384.067
DeepSeek V4.1 Flash	DeepInfra	1,000,000	$0.14	$0.42	—	0.698	164.074
DeepSeek V4.1 Flash	DigitalOcean	1,000,000	$0.30	$1.20	—	0.821	81.897
DeepSeek V4.1 Flash	GMI	1,000,000	$0.18	$0.72	—	1.813	140.702
DeepSeek V4.1 Flash	Makora	1,048,576	$0.26	$1.17	—	0.662	190.722
DeepSeek V4.1 Flash	Modal	1,000,000	$0.30	$1.20	—	0.559	157.789
DeepSeek V4.1 Flash	Nebius	1,048,576	$0.30	$1.20	—	1.334	257.082
DeepSeek V4.1 Flash	Novita	524,288	$0.30	$1.20	—	1.228	136.759
DeepSeek V4.1 Flash	Parasail	1,048,576	$0.30	$1.20	—	0.63	239.17
DeepSeek V4.1 Flash	Scaleway	262,144	$0.70	$2.09	—	1.41	201.347
DeepSeek V4.1 Flash	SiliconFlow	1,000,000	$0.15	$0.60	—	1.596	154.517
DeepSeek V4.1 Flash	Together AI	1,048,576	$0.30	$1.20	—	0.311	373.034
DeepSeek V4.1 Flash	Wafer	1,048,576	$0.30	$1.20	—	0.563	320.504
DeepSeek V4.1 Flash 1017	Baseten	1,000,000	$0.30	$1.20	393,216	0.355	166.806
DeepSeek V4.1 Flash 1017	Baseten (FAST)	1,048,576	$0.60	$2.40	393,216	0.182	1,021.729
DeepSeek V4.1 Flash 1017	Bitdeer AI	1,000,000	$0.144	$0.576	393,216	1.651	294.279
DeepSeek V4.1 Flash 1017	DeepInfra	1,000,000	$0.098	$0.294	393,216	0.701	92.086
DeepSeek V4.1 Flash 1017	DeepSeek	1,000,000	$0.30	$1.20	393,216	1.277	173.207
DeepSeek V4.1 Flash 1017	DigitalOcean	1,000,000	$0.30	$1.20	393,216	0.503	253.956
DeepSeek V4.1 Flash 1017	Fireworks	1,000,000	$0.30	$1.20	393,216	1.338	214.831
DeepSeek V4.1 Flash 1017	GMI	1,048,575	$0.18	$0.72	393,216	1.448	363.705
DeepSeek V4.1 Flash 1017	Inco (FAST)	1,048,576	$0.60	$2.40	393,216	1.317	1,075.275
DeepSeek V4.1 Flash 1017	LithosAI	1,048,576	$0.15	$0.60	393,216	0.962	248.975
DeepSeek V4.1 Flash 1017	LithosAI (FAST)	1,048,576	$0.25	$1.00	393,216	1.073	393.111
DeepSeek V4.1 Flash 1017	LithosAI (ULTRA CHAT)	1,048,576	$0.35	$1.40	393,216	—	—
DeepSeek V4.1 Flash 1017	LithosAI (ULTRA)	1,048,576	$0.35	$1.40	393,216	0.996	430.879
DeepSeek V4.1 Flash 1017	Makora	1,048,576	$0.15	$1.20	393,216	0.94	211.519
DeepSeek V4.1 Flash 1017	Modal	1,048,576	$0.25	$1.00	393,216	0.601	233.592
DeepSeek V4.1 Flash 1017	Nebius	1,048,576	$0.30	$1.20	393,216	1.33	391.095
DeepSeek V4.1 Flash 1017	Novita	1,048,576	$0.30	$1.20	393,216	1.618	149.885
DeepSeek V4.1 Flash 1017	Parasail	1,000,000	$0.30	$1.20	393,216	0.623	243.147
DeepSeek V4.1 Flash 1017	PrimaLabs	1,048,576	$0.24	$0.96	393,216	0.803	225.044
DeepSeek V4.1 Flash 1017	SiliconFlow	1,048,576	$0.15	$0.60	393,216	1.162	231.297
DeepSeek V4.1 Flash 1017	Systalyze	1,048,576	$0.30	$1.20	393,216	0.988	845.4
DeepSeek V4.1 Flash 1017	Together AI	1,048,576	$0.30	$1.20	393,216	0.31	382.936
DeepSeek V4.1 Flash 1017	Wafer	1,048,576	$0.15	$0.60	393,216	0.618	205.721
Gemini 3.1 Pro	Google (AI Studio)	1,000,000	$2.00	$12.00	65,536	22.733	143.159
Gemini 3.1 Pro	Google (Vertex)	1,000,000	$2.00	$12.00	65,536	27.154	94.583
Gemini 3.8 Flash	Google AI Studio	1,000,000	$0.75	$3.75	65,536	27.1	136.644
Gemini 4 Argon	Google (Vertex)	1,000,000	$4.00	$20.00	65,536	44.081	83.574
Gemini 4 Argon	Google AI Studio	1,000,000	$4.00	$20.00	65,536	47.152	87.964
Gemini 4 Argon	Google AI Studio (1M output)	1,000,000	$4.00	$20.00	65,536	—	—
Gemini 4 Carbon	Google	1,000,000	$4.00	$20.00	—	156.309	94.625
GLM-5.2	Z.ai	1,000,000	$1.40	$4.40	—	1.777	94.073
GLM-5.3	Baseten	1,048,576	$1.40	$4.40	131,072	2.822	106.426
GLM-5.3	Baseten (FAST)	1,000,000	$2.10	$6.60	131,072	0.795	276.567
GLM-5.3	Bitdeer AI	1,000,000	$0.63	$1.98	131,072	1.588	156.031
GLM-5.3	Crusoe (FP4)	786,432	$1.40	$4.40	131,072	0.723	269.762
GLM-5.3	Databricks	1,000,000	$1.40	$4.40	131,072	0.925	154.356
GLM-5.3	DeepInfra	1,048,575	$0.5625	$2.50	131,072	1.01	129.923
GLM-5.3	DigitalOcean	1,048,576	$1.40	$4.40	131,072	0.837	58.931
GLM-5.3	FriendliAI	1,000,000	$1.26	$3.96	131,072	0.793	139.209
GLM-5.3	GMI	1,000,000	$0.70	$2.20	131,072	2.735	125.736
GLM-5.3	Inco	163,840	$1.40	$4.40	131,072	1.539	96.309
GLM-5.3	Inco (FAST)	1,000,000	$2.80	$8.80	131,072	0.913	482.734
GLM-5.3	LithosAI (FP4 ULTRA CHAT)	1,048,576	$2.10	$6.60	131,072	—	—
GLM-5.3	LithosAI (FP4)	1,048,576	$1.05	$3.30	131,072	0.91	373.189
GLM-5.3	Makora	1,048,576	$0.14	$4.40	131,072	1.309	121.54
GLM-5.3	Modal	1,000,000	$1.40	$4.40	131,072	0.682	203.893
GLM-5.3	Modular	1,048,576	$1.40	$4.40	131,072	0.634	187.466
GLM-5.3	Multiverse Computing	1,000,000	$0.20	$0.64	131,072	1.284	214.631
GLM-5.3	Nebius (FP4)	1,000,000	$1.40	$4.40	131,072	1.492	238.818
GLM-5.3	Novita	1,000,000	$1.40	$4.40	131,072	1.938	90.783
GLM-5.3	Parasail	1,000,000	$1.40	$4.40	131,072	0.873	217.437
GLM-5.3	Scaleway	262,144	$2.09	$6.38	131,072	1.268	265.356
GLM-5.3	Supernova (FP4, FAST)	1,000,000	$2.80	$8.80	131,072	1.386	561.789
GLM-5.3	Supernova (FP4)	1,000,000	$0.63	$1.98	131,072	1.814	207.47
GLM-5.3	Systalyze	1,000,000	$0.50	$1.00	131,072	0.947	688.217
GLM-5.3	Together AI	1,048,575	$1.40	$4.40	131,072	0.597	188.682
GLM-5.3	Wafer	1,000,000	$1.40	$4.40	131,072	0.891	157.852
GLM-5.3	Zai	1,000,000	$1.40	$4.40	131,072	1.477	88.436
GLM-5.3-Flash	Z.ai	1,000,000	$0.15	$0.50	—	1.546	75.174
GPT-5.4 Pro	OpenAI	1,050,000	$30.00	$180.00	128,000	—	—
GPT-5.5	Amazon	272,000	$5.50	$33.00	128,000	51.632	142.511
GPT-5.5	OpenAI	1,050,000	$5.00	$30.00	128,000	80.223	95.007
GPT-5.5 Pro	OpenAI	922,000	$30.00	$180.00	128,000	—	—
GPT-5.6 Sol	Amazon	1,000,000	$4.00	$20.00	128,000	125.777	120.006
GPT-5.6 Sol	Azure	524,288	$4.00	$20.00	128,000	89.145	125.307
GPT-5.6 Sol	OpenAI	1,000,000	$4.00	$20.00	128,000	115.676	126.629
GPT-5.6 Terra	OpenAI	1,050,000	$2.00	$12.00	128,000	—	—
GPT-5.6 Terra	OpenAI	1,050,000	$4.00	$18.00	128,000	—	—
GPT-6 Astra	Amazon	524,288	$10.00	$50.00	128,000	302.232	89.577
GPT-6 Astra	Azure	524,288	$10.00	$50.00	128,000	349.1	69.338
GPT-6 Astra	OpenAI	1,000,000	$10.00	$50.00	128,000	365.127	70.564
GPT-6 Astra	OpenAI (Fast)	1,000,000	$20.00	$100.00	128,000	—	—
GPT-6 Astra	OpenAI (Flex)	1,000,000	$5.00	$25.00	128,000	322.971	53.896
GPT-6 Astra	OpenAI (Ultrafast)	1,000,000	$60.00	$300.00	128,000	—	—
GPT-6 Luna	OpenAI	1,050,000	$0.10	$0.50	128,000	—	—
GPT-6 Luna	OpenAI	1,050,000	$0.20	$0.75	128,000	—	—
GPT-6.1 Sol	Azure	1,048,576	$2.00	$10.00	128,000	326.239	68.704
GPT-6.1 Sol	OpenAI	1,000,000	$2.00	$10.00	128,000	315.664	88.967
GPT-6.1 Sol	OpenAI (Fast)	1,000,000	$4.00	$20.00	128,000	273.685	111.975
GPT-6.1 Sol	OpenAI (Flex)	1,000,000	$1.00	$5.00	128,000	285.681	99.892
GPT-6.1 Sol	OpenAI (Ultrafast)	1,000,000	$12.00	$60.00	128,000	64.137	342.946
GPT-6.1 Terra	OpenAI	1,050,000	$1.00	$5.00	128,000	—	—
GPT-6.1 Terra	OpenAI	1,050,000	$2.00	$7.50	128,000	—	—
Grok 4.7	SpaceXAI	500,000	$2.00	$6.00	Context-limited	111.581	87.436
Inkling	Thinking Machines	—	$1.00	$4.05	—	—	—
Kimi K2.6	Moonshot AI	256,000	$0.95	$4.00	—	2.834	40.714
Kimi K3	Amazon	1,000,000	$3.00	$15.00	1,048,576	1.421	216.139
Kimi K3	Baseten	1,048,576	$3.00	$15.00	1,048,576	1.302	90.056
Kimi K3	Databricks	204,800	$3.00	$15.00	1,048,576	0.87	115.162
Kimi K3	DigitalOcean	1,048,576	$3.00	$15.00	1,048,576	1.224	30.426
Kimi K3	Fireworks	1,048,576	$3.00	$15.00	1,048,576	0.961	113.916
Kimi K3	Inco	1,000,000	$3.00	$15.00	1,048,576	1.778	175.272
Kimi K3	Inco (FAST)	1,048,576	$6.00	$30.00	1,048,576	1.659	481.694
Kimi K3	Kimi	1,048,576	$3.00	$15.00	1,048,576	5.977	64.916
Kimi K3	LithosAI (ULTRA CHAT)	1,048,576	$5.60	$28.00	1,048,576	—	—
Kimi K3	Makora	1,048,576	$0.65	$15.00	1,048,576	1.133	84.476
Kimi K3	Modal	1,048,576	$3.00	$15.00	1,048,576	0.933	143.314
Kimi K3	Nebius	1,048,576	$3.00	$15.00	1,048,576	1.047	212.032
Kimi K3	Parasail	1,048,576	$3.00	$15.00	1,048,576	1.187	165.949
Kimi K3	Together AI	1,048,576	$3.00	$15.00	1,048,576	1.319	43.821
Kimi K3.1 Pro	Kimi	1,000,000	$2.00	$10.00	1,048,576	3.427	31.249
Kimi K3.1 Pro	Kimi High Speed	1,000,000	$4.00	$20.00	1,048,576	2.683	76.93
MiMo-V2.6-Pro	Xiaomi	—	$0.435	$0.87	—	—	—
MiniMax M3	MiniMax	1,000,000	$0.30	$1.20	—	0.699	119.672
MiniMax M3.1 Flash	DigitalOcean	1,048,576	$0.15	$0.60	524,288	0.634	325.756
MiniMax M3.1 Flash	Fireworks	1,000,000	$0.15	$0.60	524,288	0.929	160.733
MiniMax M3.1 Flash	GMI	1,000,000	$0.1425	$0.57	524,288	2.268	219.719
MiniMax M3.1 Flash	Inco	262,144	$0.30	$1.20	524,288	1.403	812.597
MiniMax M3.1 Flash	Inco	1,000,000	$0.15	$0.60	524,288	1.843	89.294
MiniMax M3.1 Flash	MiniMax	1,000,000	$0.15	$0.60	524,288	0.904	136.2
MiniMax M3.1 Flash	Modular	1,000,000	$0.15	$0.60	524,288	0.893	209.151
MiniMax M3.1 Flash	Nebius	1,000,000	$0.15	$0.60	524,288	1.723	114.21
MiniMax M3.1 Flash	Novita	1,000,000	$0.15	$0.60	524,288	1.613	127.737
MiniMax M3.1 Flash	Parasail	1,000,000	$0.15	$0.60	524,288	0.849	190.503
MiniMax M3.1 Flash	SambaNova	1,048,576	$0.30	$1.20	524,288	2.213	159.122
MiniMax M3.1 Flash	SiliconFlow	1,000,000	$0.15	$0.60	524,288	2.739	151.741
MiniMax M3.1 Flash	Together AI	1,048,576	$0.15	$0.60	524,288	0.584	140.861
Muse Spark 1.1	Meta	1,048,576	$1.25	$4.25	—	12.875	294.835
Muse Spark 1.2	Meta	1,048,576	$1.25	$4.25	—	12.279	277.574
Muse Spark 1.3	Meta	1,048,576	$1.25	$4.25	—	45.675	195.207
Muse Wonder 1.0	Meta	1,048,576	$4.00	$20.00	—	—	—

Best models per task

Fastest and most affordable models

Model comparisons