Language models

The best LLM at every budget.

We score 149 models from 19 providers on up to 17 benchmarks and price them per million tokens. Every model is in the table at the bottom.

Weekly notes on what changedSubscribe

Best quality, any price

Gemini 4 Argon

Provisional, 2 benchmarks

114quality

$8.00 / 1M

The strongest model we can price today.

Best value

GLM-5.3 Flash (Thinking)

Provisional, 3 benchmarks

85quality

$0.08 / 1M

75% of the top score for 1% of the price.

The best model under each price

Each row is the strongest model under that blended price per million tokens.

BudgetModelQuality$ / 1M

Under $0.15

GLM-5.3 Flash (Thinking)Best value · Provisional, 3 benchmarks85$0.08

Under $0.25

DeepSeek V4 Flash Thinking91$0.16

Under $2

DeepSeek V4.1 Flash ThinkingProvisional, 4 benchmarks94$0.39

Under $3

Gemini 3.7 FlashProvisional, 3 benchmarks98$3.00

Under $5

Gemini 3.8 Flash102$3.30

Under $25

Gemini 4 ArgonBest quality · Provisional, 2 benchmarks114$8.00

How sure are we?

The dot is our estimate, the bar the range the score most likely falls in. Where bars overlap, the order is within the noise: compare those models on price, not on rank.

Model

Gemini 4 ArgonProvisional, 2 benchmarks
Claude Opus 5.5Provisional, 4 benchmarks
Claude Mythos 5Provisional, 2 benchmarks
Claude Mythos (Preview)
GPT-6 Astra (Thinking)
Claude Fable 5.1
GPT-6.1 Sol (Thinking)Provisional, 4 benchmarks
Claude Fable 5 (Thinking)
Muse Spark 1.3Provisional, 3 benchmarks
Claude Opus 5 (Thinking)
Gemini 3.8 Flash
Muse Spark 1.1Provisional, 4 benchmarks

Price against quality, one benchmark at a time.

Each dot is a model. The dashed line joins the ones nothing beats on both price and score; everything below it has a cheaper or stronger alternative. Switch the benchmark to see who leads at coding, maths or agent tasks.

198 models compared50 = baseline, 100 = top 5%, 100+ = frontier; higher coverage yields higher confidence

AnthropicCohereDeepSeekGoogleMetaMicrosoftMiniMaxMistralMoonshotNVIDIAOpenAIQwenThinking MachinesxAIZhipu

Every model we track.

314 variants, one row each. Click a column to sort. To estimate a monthly bill for your own mix of input and output tokens, use the pricing calculator.


Gemini 4 ArgonGoogle · Provisional, 2 benchmarks	114	102	1520	1.0M	$2	$10
Anthropic Claude Opus 5.5Anthropic · Provisional, 4 benchmarks	114	98	1509	1.0M	$4	$20
Claude Mythos 5Anthropic · Provisional, 2 benchmarks	113	—	—	—	—	—
Claude Mythos PreviewAnthropic · Preview	112	—	—	—	—	—
GPT-6 Astra ThinkingOpenAI	108	87	1476	1.1M	$10	$50
Claude Fable 5.1Anthropic	108	87	1504	1.0M	$10	$50
GPT-6.1 Sol ThinkingOpenAI · Provisional, 4 benchmarks	105	94	1481	1.1M	$2	$10
Claude Fable 5 ThinkingAnthropic	104	84	1504	1.0M	$10	$50
Meta Muse Spark 1.3Meta · Provisional, 3 benchmarks	104	—	1495	—	—	—
Anthropic Claude Opus 5 ThinkingAnthropic	103	87	1509	1.0M	$5	$25
Gemini 3.8 FlashGoogle	102	103	1496	1.0M	$0.83	$4.13
Meta Muse Spark 1.1Meta · Provisional, 4 benchmarks	100	—	1495	—	—	—
GPT-5.6 Sol ThinkingOpenAI	99	85	1485	1.1M	$4	$20
Gemini 3.7 FlashGoogle · Provisional, 3 benchmarks	98	103	1490	1.0M	$0.75	$3.75
Anthropic Claude Opus 4 4.8 ThinkingAnthropic	98	82	1503	200K	$5	$25
Meta Muse Spark 1.2Meta · Provisional, 3 benchmarks	97	—	1495	—	—	—
Anthropic Claude Opus 4 4.7 ThinkingAnthropic	97	81	1503	1.0M	$5	$25
Qwen3 Max Qwen 3.8 MaxQwen · Provisional, 2 benchmarks	97	100	1482	262K	$0.78	$3.90
Anthropic Claude Sonnet 5.5Anthropic	96	86	1477	1.0M	$2	$10
Gemini 3 Pro 3.1Google	95	84	1487	1.0M	$2	$12
GPT-6 Sol ThinkingOpenAI · Provisional, 4 benchmarks	95	85	1458	1.1M	$2	$10
GPT-5.3 CodexOpenAI · Provisional, 3 benchmarks	94	83	1482	400K	$1.75	$14
GPT-5.5 ThinkingOpenAI	94	91	1482	400K	$1.25	$10
Anthropic Claude Opus 4 4.6 ThinkingAnthropic	94	79	1503	1.0M	$5	$25
DeepSeek V4 Pro ThinkingDeepSeek · Open weights	94	105	1465	1.0M	$1.60	$3.20
DeepSeek V4.1 Flash ThinkingDeepSeek · Open weights · Provisional, 4 benchmarks	94	133	1476	1.0M	$0.04	$0.50
Qwen 3.8 2.4T A95B ThinkingQwen · Open weights · Provisional, 2 benchmarks	93	93	—	1.0M	$2	$6
Moonshot Kimi K2 K2.6 ThinkingMoonshot · Open weights	93	102	1461	131K	$0.57	$2.30
GLM-5.3 ThinkingZhipu · Provisional, 4 benchmarks	92	105	1481	1.0M	$0.07	$1.40
Moonshot Kimi K3 ThinkingMoonshot · Provisional, 5 benchmarks	92	80	1488	1.0M	$0.19	$16
Gemini 3.5 FlashGoogle	92	89	1478	1.0M	$1.50	$9
GPT-5.4 ThinkingOpenAI	91	79	1482	1.1M	$2.50	$15
DeepSeek V4 Flash ThinkingDeepSeek · Open weights	91	138	1465	1.0M	$0.09	$0.18
xAI Grok 4.6xAI	91	90	1475	500K	$2	$6
Meta Muse Spark ThinkingMeta	90	—	1495	—	—	—
Qwen3 Max Qwen 3.7 Max PreviewQwen · Preview	90	94	1482	262K	$0.78	$3.90
Gemini 3 Pro 3.0Google	90	80	1487	1.0M	$2	$12
Anthropic Claude Opus 4 4.5 ThinkingAnthropic	90	75	1503	200K	$5	$25
Qwen 3.5 122B A10B Non-ThinkingQwen · Open weights	89	100	1416	262K	$0.26	$2.08
xAI Grok 4.5xAI	89	89	1475	2.0M	$2	$6
xAI Grok 4 4.20 Beta1 (Reasoning)xAI	89	98	1475	1.0M	$1.25	$2.50
Anthropic Claude Sonnet 4 4.6 ThinkingAnthropic	89	77	1472	200K	$3	$15
Qwen 3.5 397B A17B ThinkingQwen · Open weights	89	96	1441	262K	$0.50	$3.60
Gemini 3.6 FlashGoogle	88	93	1483	1.0M	$0.75	$3.75
Anthropic Claude Sonnet 5 ThinkingAnthropic	88	78	1477	1.0M	$2	$10
GPT-5.2 ThinkingOpenAI	88	77	1482	400K	$1.75	$14
GPT-5.6 Terra ThinkingOpenAI	88	77	1467	1.1M	$2	$12
GLM-5.1 ThinkingZhipu · Open weights	88	89	1465	205K	$1.40	$4.40
Qwen3 Max Qwen 3.6 Max PreviewQwen · Preview	87	87	1482	262K	$1.03	$6.16
GLM-5.2 ThinkingZhipu · Open weights	87	87	1475	1.0M	<$0.01	$6.80
Google Gemma 4 31B IT ThinkingGoogle · Open weights	87	102	1452	262K	$0.20	$0.40
Qwen3 MaxQwen	87	91	1482	262K	$0.78	$3.90
Anthropic Claude Opus 4 4.6 Non-thinkingAnthropic	86	72	1503	200K	$5	$25
xAI Grok 4.7xAI	86	74	1475	256K	$3	$15
GLM-5.3 Flash ThinkingZhipu · Open weights	85	133	1477	1.0M	$0.03	$0.10
Qwen 3.6 Plus ThinkingQwen	85	96	1443	1.0M	$0.33	$1.95
xAI Grok 4.2xAI	85	73	1475	256K	$3	$15
Mistral Medium 3.5 ThinkingMistral · Open weights	85	84	1426	262K	$1.50	$7.50
GLM-5 ThinkingZhipu · Open weights	84	92	1457	205K	$1	$3.20
Moonshot Kimi K2 K2.5 ThinkingMoonshot · Open weights	84	94	1461	262K	$0.50	$2.50
MiniMax M3 ThinkingMiniMax · Open weights	83	96	1439	1.0M	$0.30	$1.20
GPT-5.6 Luna ThinkingOpenAI	83	95	1453	1.1M	$0.20	$1.20
Gemini 3 Flash 3.0Google	83	92	1473	—	$0.50	$3
Qwen 3.5 122B A10B ThinkingQwen · Open weights	83	94	1416	262K	$0.26	$2.08
GPT-5.0OpenAI	82	80	1482	400K	$1.25	$10
Qwen 3.6 27B ThinkingQwen · Open weights	82	92	—	262K	$0.45	$2.70
DeepSeek v3 3.2 ThinkingDeepSeek · Open weights	82	96	1425	164K	$0.26	$0.38
Qwen 3.5 27B ThinkingQwen · Open weights	82	92	1409	262K	$0.26	$2.60
DeepSeek v3 V3.2 Exp ChatDeepSeek · Open weights	82	95	1425	164K	$0.27	$0.41
Claude Haiku 5.5 ThinkingAnthropic	81	93	1425	1.0M	$0.10	$0.50
MiniMax M2 M2.7MiniMax	81	93	1415	205K	$0.26	$1.02
Qwen 3.8 27B ThinkingQwen · Open weights	81	93	1436	1.0M	$0.15	$1.88
xAI Grok 4xAI	81	70	1475	256K	$3	$15
Grok 4 Fast ThinkingxAI	81	93	1430	2.0M	$0.20	$0.50
Google Gemma 4 26B A4B IT ThinkingGoogle · Open weights	81	94	1438	262K	$0.10	$0.28
Anthropic Claude Sonnet 4 4.5 ThinkingAnthropic	81	70	1472	1.0M	$3	$15
GPT-5 Codex Non-thinkingOpenAI	81	83	—	400K	$0.63	$5
NVIDIA Nemotron 3 Ultra 550B-A55BNVIDIA · Open weights	81	91	1427	262K	$0.60	$2.40
OpenAI o3OpenAI	80	79	1432	200K	$2	$8
Moonshot Kimi K2 ThinkingMoonshot · Open weights	80	90	1461	262K	$0.60	$2.50
Qwen 3.5 9B ThinkingQwen · Open weights	80	92	—	262K	$0.17	$0.25
GPT-6 Luna ThinkingOpenAI	79	90	1441	1.1M	$0.10	$0.50
Inkling SmallThinking Machines · Open weights	79	84	1442	524K	$1	$4.05
GPT-5.1 ThinkingOpenAI	78	76	1482	400K	$1.25	$10
Anthropic Claude Opus 4 4.1 ThinkingAnthropic	78	58	1503	200K	$15	$75
Grok 4 Fast Non-ThinkingxAI	78	89	1430	2.0M	$0.20	$0.50
Inkling ThinkingThinking Machines · Open weights	78	83	1442	524K	$1	$4.05
Qwen3 Next 80B A3B ThinkingQwen · Open weights	78	88	—	262K	$0.15	$1.20
Qwen 3 30B A3B ThinkingQwen · Open weights	78	88	1382	82K	$0.20	$2.40
Qwen 3.6 35B-A3B ThinkingQwen · Open weights	78	88	—	262K	$0.10	$0.90
GPT-5 Mini Thinking (5.4)OpenAI	78	88	1447	400K	$0.25	$2
Qwen 3.5 35B A3B ThinkingQwen · Open weights	77	87	1394	262K	$0.14	$1
GPT-5.2OpenAI	77	67	1482	400K	$1.75	$14
Anthropic Claude Opus 4 4.0 ThinkingAnthropic	77	56	1503	200K	$15	$75
MiniMax M2 M2.5MiniMax	77	87	1415	205K	$0.26	$1.02
OpenAI o3 Pro (Extended Reasoning)OpenAI	77	56	1432	200K	$20	$80
Google Gemini 2.5 ProGoogle	76	75	1445	—	$1.25	$10
GLM-4 4.7Zhipu · Open weights	76	86	1441	205K	$0.60	$2.20
Gemini 3.5 Flash LiteGoogle	76	85	1456	1.0M	$0.30	$2.50
Gemini 3 Flash PreviewGoogle · Preview	76	85	1473	—	$0.50	$3
Anthropic Claude Opus 4 4.5 Non-thinkingAnthropic	76	62	1503	200K	$5	$25
Qwen 3 14B ThinkingQwen · Open weights	76	86	—	41K	$0.12	$0.24
Mistral Large 4Mistral · Open weights	75	—	1432	—	—	—
GPT-5 Mini Thinking (5.0)OpenAI	75	85	1447	400K	$0.25	$2
DeepSeek V4 ProDeepSeek · Open weights	75	84	1465	1.0M	$1.60	$3.20
OpenAI o4 Mini o4-miniOpenAI	75	78	1391	200K	$1.10	$4.40
DeepSeek R1 DeepSeek-R1-0528DeepSeek · Open weights	75	84	1421	164K	$0.50	$2.15
Moonshot Kimi K2 K2.7 CodeMoonshot · Open weights	75	83	1461	262K	$0.60	$3.50
DeepSeek v3 3.2DeepSeek · Open weights	75	85	1425	164K	$0.26	$0.38
Microsoft Phi-4 Reasoning PlusMicrosoft · Open weights	75	88	1256	16K	$0.07	$0.14
GLM-4 4.5 ThinkingZhipu · Open weights	74	83	1441	131K	$0.60	$2.20
GPT-5.2 CodexOpenAI	74	64	1482	400K	$1.75	$14
DeepSeek V4 FlashDeepSeek · Open weights	73	83	1465	1.0M	$0.09	$0.18
Qwen 3.5 4B ThinkingQwen · Open weights	73	—	—	—	—	—
Gemini 3.1 Flash Lite LatestGoogle · Preview	73	81	1433	—	$0.25	$1.50
Anthropic Claude Opus 4 4.0 Non-thinkingAnthropic	73	53	1503	200K	$15	$75
xAI Grok 3 ThinkingxAI	73	62	1411	131K	$3	$15
Qwen 3 8B ThinkingQwen · Open weights	73	82	—	131K	$0.12	$0.46
GPT-5 Nano Thinking (5.4)OpenAI	72	82	1401	400K	$0.05	$0.40
Claude Haiku 4.5 ThinkingAnthropic	72	75	1415	200K	$1	$5
Qwen 3 235B A22B thinking-2507Qwen · Open weights	72	80	1422	128K	$0.45	$3.50
Anthropic Claude Sonnet 4 4.0 ThinkingAnthropic	72	62	1472	200K	$3	$15
xAI Grok 4.3xAI	72	80	1475	1.0M	$1.25	$2.50
Moonshot Kimi K2 0905 PreviewMoonshot · Open weights · Preview	72	80	1461	262K	$0.60	$2.50
GLM-4 4.7 FlashZhipu · Open weights	72	81	1441	200K	$0.06	$0.40
Microsoft Phi-4 ReasoningMicrosoft · Open weights	72	84	1256	16K	$0.07	$0.14
DeepSeek R1 ThinkingDeepSeek · Open weights	71	79	1421	64K	$0.70	$2.50
DeepSeek v3 3.1DeepSeek · Open weights	71	79	1425	164K	$0.25	$0.95
Grok 3 MinixAI	71	78	—	131K	$0.30	$0.50
Google Gemma 4 12B IT ThinkingGoogle · Open weights	71	—	—	—	—	—
GPT-OSS 120B ThinkingOpenAI · Open weights	71	82	1352	131K	$0.03	$0.14
MiniMax M2 M2.1MiniMax	71	78	1415	205K	$0.30	$1.20
OpenAI o1OpenAI	71	52	1402	200K	$15	$60
Qwen 3 235B A22B ThinkingQwen · Open weights	71	78	1422	131K	$0.46	$1.82
Qwen 3 32B ThinkingQwen · Open weights	71	79	1347	131K	$0.08	$0.28
Anthropic Claude 3 Sonnet 3.7 ThinkingAnthropic	70	60	1388	200K	$3	$15
GPT-OSS 20B ThinkingOpenAI · Open weights	70	80	1317	131K	$0.02	$0.11
QwQ 32B ThinkingQwen · Open weights	70	78	1336	33K	$0.20	$0.20
Anthropic Claude Sonnet 4 4.5 Non-thinkingAnthropic	69	59	1472	1.0M	$3	$15
GPT-5.1OpenAI	69	68	1482	400K	$1.25	$10
Qwen3-4B (Thinking) ThinkingQwen · Open weights	69	—	—	41K	<$0.01	<$0.01
Qwen 3 235B A22B non-thinking-2507Qwen · Open weights	69	76	1422	262K	$0.14	$0.80
Moonshot Kimi K2 0711 PreviewMoonshot · Open weights · Preview	69	76	1461	131K	$0.57	$2.30
MiniMax M1 Reasoning 80KMiniMax	69	75	1364	1.0M	$0.55	$2.20
Anthropic Claude Sonnet 4 4.0 Non-thinkingAnthropic	69	59	1472	200K	$3	$15
Google Gemini 2.0 Pro Non-thinkingGoogle	68	—	—	—	—	—
Google Gemini 2.5 Flash ThinkingGoogle	68	74	1409	1.0M	$0.30	$2.50
MistralAI Magistral Small ThinkingMistral	68	73	—	40K	$0.50	$1.50
Google Gemma 4 E4B IT ThinkingGoogle · Open weights	68	—	—	—	—	—
OpenAI o1 PreviewOpenAI · Preview	68	49	1402	200K	$15	$60
NVIDIA Nemotron 3 Super 120B-A12BNVIDIA · Open weights	67	73	1360	1.0M	$0.30	$2.40
MiniMax M1 Reasoning 40KMiniMax	67	73	1364	1.0M	$0.55	$2.20
Anthropic Claude Opus 4 4.1 Non-thinkingAnthropic	67	48	1503	200K	$15	$75
GLM-4 4.5Zhipu · Open weights	67	72	1441	131K	$0.60	$2.20
Qwen 3 30B A3B Thinking (2507)Qwen · Open weights	67	72	1382	41K	$0.12	$0.50
GLM-4.5 Air ThinkingZhipu	66	71	1373	131K	$0.13	$0.85
OpenAI o3 Mini o3-miniOpenAI	66	67	1364	200K	$1.10	$4.40
MistralAI Magistral Medium ThinkingMistral	66	67	1305	41K	$2	$5
Claude Haiku 4.5 Non-thinkingAnthropic	65	66	1415	200K	$1	$5
Google Gemini 2.5 Flash-Lite LatestGoogle	65	71	1379	—	$0.10	$0.40
Anthropic Claude 3 Sonnet 3.7Anthropic	65	55	1388	200K	$3	$15
DeepSeek v3 v3-0324 (Non-thinking)DeepSeek · Open weights	65	69	1425	164K	$0.29	$1.14
Google Gemini 2.5 Flash Non-ThinkingGoogle	64	68	1409	1.0M	$0.30	$2.50
OpenAI GPT-4.5 Non-thinkingOpenAI	64	—	1445	—	—	—
Qwen 3 14B Non-thinkingQwen · Open weights	64	69	—	41K	$0.12	$0.24
Qwen 3 32B Non-thinkingQwen · Open weights	63	68	1347	131K	$0.08	$0.28
OpenAI GPT-4.1 Non-thinkingOpenAI	63	55	1415	—	$3	$12
DeepSeek v3 (Non-thinking)DeepSeek · Open weights	63	67	1425	164K	$0.26	$1.03
Qwen 3 235B A22B Non-ThinkingQwen · Open weights	63	66	1422	131K	$0.46	$1.82
Phi-4 Mini Flash ReasoningMicrosoft · Open weights	63	—	—	128K	—	—
Qwen 3 30B A3B Non-thinkingQwen · Open weights	62	66	1382	41K	$0.12	$0.50
Qwen 3 Coder 480B A35B Instruct Non-thinkingQwen · Open weights	62	65	1388	262K	$0.30	$1
OpenAI o1 ProOpenAI	62	44	1402	200K	$15	$60
Gemini 2.0 Flash-Lite Non-thinkingGoogle	62	66	1354	1.0M	$0.07	$0.30
Phi-4 Mini ReasoningMicrosoft · Open weights	62	66	—	128K	$0.08	$0.35
Google Gemma 4 E2B IT ThinkingGoogle · Open weights	62	—	—	—	—	—
MistralAI Mistral Large 3Mistral · Open weights	61	64	1414	262K	$0.50	$1.50
Google Gemini 2.0 FlashGoogle	61	65	1360	1.0M	$0.10	$0.40
Qwen3-1.7B (Thinking) ThinkingQwen · Open weights	61	—	—	—	—	—
Anthropic Claude 3 Sonnet 3.5 (October 2024)Anthropic	61	51	1388	200K	$3	$15
MiniMax M2 ReasoningMiniMax	60	63	1415	205K	$0.26	$1.02
GPT-OSS 120B Non-thinkingOpenAI · Open weights	60	67	1352	131K	$0.03	$0.14
Moonshot Kimi K2 BaseMoonshot · Open weights	60	62	1461	131K	$0.57	$2.30
GLM-4 4.6Zhipu · Open weights	60	61	1441	205K	$0.50	$2
Anthropic Claude 3 Sonnet 3.5 (June 2024)Anthropic	59	50	1388	200K	$3	$15
GPT-OSS 20B Non-thinkingOpenAI · Open weights	59	65	1317	131K	$0.02	$0.11
Moonshot Kimi K2 InstructMoonshot · Open weights	59	60	1461	131K	$0.57	$2.30
Meta Llama 4 Maverick Base (Non-thinking)Meta · Open weights	59	60	1327	524K	$0.19	$0.65
MistralAI Small 3Mistral · Open weights	58	61	1356	256K	$0.07	$0.20
Qwen 3 8B Non-thinkingQwen · Open weights	58	61	—	131K	$0.12	$0.46
Qwen3-4B (Thinking) Non-thinkingQwen · Open weights	58	—	—	41K	<$0.01	<$0.01
DeepSeek v3 BaseDeepSeek · Open weights	58	60	1425	164K	$0.26	$1.03
OpenAI GPT-4.1 Mini Non-thinkingOpenAI	58	60	1383	—	$0.80	$3.20
Meta Llama 4 Scout Non-thinkingMeta · Open weights	58	61	1321	1.3M	$0.10	$0.30
Google Gemini 2.0 Flash ReasoningGoogle	58	61	1360	1.0M	$0.10	$0.40
OpenAI o1 Mini ThinkingOpenAI	58	55	1337	128K	$1.10	$4.40
Meta Llama 4 Maverick Default (Non-thinking)Meta · Open weights	58	59	1327	524K	$0.19	$0.65
Google Gemma 3 27B IT Non-thinkingGoogle · Open weights	58	60	1365	131K	$0.08	$0.35
Anthropic Claude 3.5 Haiku Non-thinkingAnthropic	57	54	1325	200K	$1	$5
Google Gemma 3 12B IT Non-thinkingGoogle · Open weights	57	62	1342	131K	$0.05	$0.15
GPT-5 Nano Thinking (5.0)OpenAI	56	58	1401	400K	$0.05	$0.40
Microsoft Phi-4 Non-ThinkingMicrosoft · Open weights	56	61	1256	16K	$0.07	$0.14
Meta Llama 3 70B 3.3Meta · Open weights	56	58	1318	131K	$0.10	$0.32
Meta Llama 3 405B 3.1Meta · Open weights	56	52	1335	131K	$4	$4
Cohere Command A Non-thinkingCohere	56	45	1354	256K	$2.50	$10
MistralAI Mistral Medium 3Mistral	54	54	1408	131K	$0.40	$2
OpenAI GPT-4o Non-thinkingOpenAI	53	47	1443	128K	$1.25	$5
Qwen3-0.6B (Thinking) ThinkingQwen · Open weights	50	—	—	—	—	—
Google Gemma 3n 4B Non-thinkingGoogle · Open weights	50	52	1318	33K	$0.06	$0.12
OpenAI GPT-4.1 Nano Non-thinkingOpenAI	50	49	1322	1.0M	$0.10	$0.40
Google Gemma 3 4B IT Non-thinkingGoogle · Open weights	50	51	1303	131K	$0.05	$0.10
OpenAI GPT-4o Mini Non-thinkingOpenAI	49	47	1318	128K	$0.07	$0.30
Qwen3-1.7B (Thinking) Non-thinkingQwen · Open weights	48	—	—	—	—	—
QwQ 32B PreviewQwen · Open weights · Preview	47	45	1336	33K	$0.20	$0.20
Google Gemma 3n 2B IT Non-thinkingGoogle · Open weights	47	—	—	8K	<$0.01	<$0.01
Phi-4 Mini Non-ThinkingMicrosoft · Open weights	46	43	—	128K	$0.08	$0.35
Meta Llama 3 8B 3.1 (Non-thinking)Meta · Open weights	44	64	1224	131K	$0.05	$0.08
Cohere Command R Non-thinkingCohere	40	33	1250	128K	$0.50	$1.50
Qwen3-0.6B (Thinking) Non-thinkingQwen · Open weights	38	—	—	—	—	—
Google Gemma 3 1B IT Non-thinkingGoogle · Open weights	33	—	—	—	—	—
Anthropic Claude Opus 4 4.7 Non-thinkingAnthropic	—	—	1503	200K	$5	$25
Anthropic Claude Opus 4 4.8 Non-thinkingAnthropic	—	—	1503	200K	$5	$25
Anthropic Claude 3 Sonnet 3.0Anthropic	—	—	1388	200K	$3	$15
Cohere Command R 08-2024Cohere	—	—	1250	128K	$0.50	$1.50
DeepSeek v3 3.1-terminusDeepSeek · Open weights	—	—	1425	164K	$0.27	$1
DeepSeek v3 3.1-terminus-thinkingDeepSeek · Open weights	—	—	1425	164K	$0.27	$1
DeepSeek v3 3.1-thinkingDeepSeek · Open weights	—	—	1425	164K	$0.25	$0.95
DeepSeek v3 V3.2 SpecialeDeepSeek · Open weights	—	—	1425	164K	$0.29	$0.43
DeepSeek v3 v3.2-expDeepSeek · Open weights	—	—	1425	164K	$0.27	$0.41
DeepSeek v3 V3.2 Exp ThinkingDeepSeek · Open weights	—	—	1425	164K	$0.27	$0.41
DeepSeek v3 V3.2 ThinkingDeepSeek · Open weights	—	—	1425	164K	$0.26	$0.38
DeepSeek V4 Flash Vision ExpDeepSeek · Open weights	—	—	1465	1.0M	$0.22	$0.65
Google Gemini 2.0 Flash Preview 01 01Google · Preview	—	—	1360	1.0M	$0.10	$0.40
Google Gemini 2.5 Flash-Lite Max Thinking 2025 06 17Google	—	—	1379	—	$0.10	$0.40
Google Gemini 2.5 Flash-Lite Max Thinking 2025 09 25Google	—	—	1379	—	$0.10	$0.40
Google Gemini 2.5 Flash-Lite Preview 06 17 ThinkingGoogle · Preview	—	—	1379	—	$0.10	$0.40
Google Gemini 2.5 Flash-Lite Preview 09 2025 No ThinkingGoogle · Preview	—	—	1379	—	$0.10	$0.40
Google Gemini 2.5 Flash Max ThinkingGoogle	—	—	1409	1.0M	$0.30	$2.50
Granite 4.2 30BIBM · Open weights	—	—	1340	—	—	—
Granite 4.2 3BIBM · Open weights	—	—	1290	—	—	—
Granite 4.2 8BIBM · Open weights	—	—	1290	—	—	—
Meta Llama 3 405B 3.1 BF16Meta · Open weights	—	—	1335	131K	$4	$4
Meta Llama 3 405B 3.1 FP8Meta · Open weights	—	—	1335	131K	$4	$4
Meta Llama 3 70B 3.0Meta · Open weights	—	—	1318	8K	$0.51	$0.74
Meta Llama 3 70B 3.1Meta · Open weights	—	—	1318	131K	$0.72	$0.72
Meta Llama 3 8B 3.0 (Non-thinking)Meta · Open weights	—	—	1224	131K	$0.05	$0.08
Phi-4 Multimodal Instruct Non-ThinkingMicrosoft · Open weights	—	—	—	—	—	—
Phi-4 Reasoning Vision 15B Non-ThinkingMicrosoft · Open weights	—	—	—	—	—	—
Phi-4 Reasoning Vision 15B ThinkingMicrosoft · Open weights	—	—	—	—	—	—
Phi-4 Mini Flash Non-ThinkingMicrosoft · Open weights	—	—	—	128K	—	—
MiniMax M1 ReasoningMiniMax	—	—	1364	1.0M	$0.55	$2.20
MistralAI Mistral Large 2402 (February 2024)Mistral · Open weights	—	—	1414	131K	$2	$6
MistralAI Mistral Large 2407 (July 2024)Mistral · Open weights	—	—	1414	131K	$2	$6
MistralAI Mistral Large 2411 (November 2024)Mistral · Open weights	—	—	1414	131K	$2	$6
MistralAI Mistral LargeMistral · Open weights	—	—	1414	131K	$2	$6
MistralAI Mistral Medium 3 2505 (May 2025)Mistral	—	—	1408	131K	$0.40	$2
MistralAI Mistral Medium 3.1Mistral	—	—	1408	131K	$0.40	$2
MistralAI Small 3 2501 (January 2025)Mistral · Open weights	—	—	1356	33K	$0.05	$0.08
MistralAI Small 3 2503 (March 2025)Mistral · Open weights	—	—	1356	128K	$0.35	$0.55
MistralAI Small 3 2506 (June 2025)Mistral · Open weights	—	—	1356	256K	$0.07	$0.20
Moonshot Kimi K2 K2.5 InstantMoonshot · Open weights	—	—	1461	131K	$0.57	$2.30
Moonshot Kimi K2 K2.6Moonshot · Open weights	—	—	1461	131K	$0.57	$2.30
Moonshot Kimi K2 Thinking TurboMoonshot · Open weights	—	—	1461	131K	$0.57	$2.30
Kimi-VL-A3B Non-ThinkingMoonshot · Open weights	—	—	—	—	—	—
Kimi-VL-A3B ThinkingMoonshot · Open weights	—	—	—	—	—	—
NVIDIA Nemotron 3 Nano Omni 30B-A3B InstructNVIDIA · Open weights	—	—	—	256K	<$0.01	<$0.01
NVIDIA Nemotron 3 Nano Omni 30B-A3B ReasoningNVIDIA · Open weights	—	—	—	256K	<$0.01	<$0.01
GPT-5.6 Luna Non-thinkingOpenAI	—	—	1453	1.1M	$0.20	$1.20
GPT-5 Nano Non-Thinking (5.0)OpenAI	—	—	1401	400K	$0.05	$0.40
GPT-5.1 CodexOpenAI	—	—	1482	400K	$1.25	$10
GPT-5.1 Codex MaxOpenAI	—	—	1482	400K	$1.25	$10
GPT-5.1 Codex MiniOpenAI	—	—	1482	400K	$0.25	$2
GPT-5.3OpenAI	—	—	1482	400K	$1.25	$10
GPT-5.4OpenAI	—	—	1482	1.1M	$2.50	$15
GPT-5.5OpenAI	—	—	1482	400K	$1.25	$10
O InternVL3.5 8B Non-ThinkingOpenGVLab · Open weights	—	—	—	—	—	—
O InternVL3.5 8B ThinkingOpenGVLab · Open weights	—	—	—	—	—	—
Qwen 3 235B A22B vl-235b-a22b-instructQwen · Open weights	—	—	1422	131K	$0.46	$1.82
Qwen 3 235B A22B vl-235b-a22b-thinkingQwen · Open weights	—	—	1422	131K	$0.46	$1.82
Qwen 3 30B A3B Non-Thinking (2507)Qwen · Open weights	—	—	1382	262K	$0.10	$0.30
Qwen 3.5 0.8B Non-ThinkingQwen · Open weights	—	—	—	—	—	—
Qwen 3.5 0.8B ThinkingQwen · Open weights	—	—	—	—	—	—
Qwen 3.5 27B Non-ThinkingQwen · Open weights	—	—	1409	262K	$0.26	$2.60
Qwen 3.5 2B Non-ThinkingQwen · Open weights	—	—	—	—	—	—
Qwen 3.5 2B ThinkingQwen · Open weights	—	—	—	—	—	—
Qwen 3.5 35B A3B Non-ThinkingQwen · Open weights	—	—	1394	262K	$0.14	$1
Qwen 3.5 4B Non-ThinkingQwen · Open weights	—	—	—	—	—	—
Qwen 3.5 9B Non-ThinkingQwen · Open weights	—	—	—	262K	$0.17	$0.25
Qwen 3.5 Flash ThinkingQwen · Open weights	—	—	1396	1.0M	$0.07	$0.26
Qwen 3.6 Flash ThinkingQwen · Open weights	—	—	—	1.0M	$0.19	$1.13
Qwen 3.7 Plus ThinkingQwen	—	—	1456	1.0M	$0.32	$1.28
Qwen 3.8 Flash Next ThinkingQwen · Open weights	—	—	—	—	—	—
Qwen3 Max Qwen 3.5 Max PreviewQwen · Preview	—	—	1482	262K	$0.78	$3.90
Qwen3 VL 235B A22B ThinkingQwen · Open weights	—	—	—	131K	$0.40	$4
Qwen3 VL 2B ThinkingQwen · Open weights	—	—	—	—	—	—
Qwen3 VL 30B A3B ThinkingQwen · Open weights	—	—	—	262K	$0.29	$1
Qwen3 VL 32B Non-ThinkingQwen · Open weights	—	—	—	131K	$0.10	$0.42
Qwen3 VL 32B ThinkingQwen · Open weights	—	—	—	131K	$0.10	$0.42
Qwen3 VL 4B ThinkingQwen · Open weights	—	—	—	—	—	—
Qwen3 VL 8B Non-ThinkingQwen · Open weights	—	—	—	262K	$0.25	$0.75
Qwen3 VL 8B ThinkingQwen · Open weights	—	—	—	131K	$0.18	$2.10
T Penguin-VL 8B Non-ThinkingTencent · Open weights	—	—	—	—	—	—
T Penguin-VL 8B ThinkingTencent · Open weights	—	—	—	—	—	—
Inkling Small (Non-thinking)Thinking Machines · Open weights	—	—	1442	524K	$1	$4.05
xAI Grok 4 4.1xAI	—	—	1475	256K	$3	$15
xAI Grok 4 4.1 ThinkingxAI	—	—	1475	256K	$3	$15
xAI Grok 4 4.20 Beta1xAI	—	—	1475	1.0M	$3	$15
xAI Grok 4 4.20 Beta1 (Non-Thinking)xAI	—	—	1475	1.0M	$1.25	$2.50
xAI Grok 4 4.20 Multi-AgentxAI	—	—	1475	2.0M	$1.25	$2.50
Grok Build 0.1xAI	—	—	—	256K	$1	$2
Grok Code Fast Non-thinkingxAI	—	—	—	256K	$0.20	$1.50
MiMo-V2.6-Flash v2.6-flashXiaomi · Open weights	—	—	1448	1.1M	$0.14	$0.28
MiMo-V2.6-Pro v2.6-proXiaomi · Open weights	—	—	1480	1.1M	$0.43	$0.87
GLM-5 5V TurboZhipu · Open weights	—	—	1457	205K	$1	$3.20