Find an AI model for the work you need done

Compare 104 models from 24 providers by RankLLMs coding, reasoning, and agent scores. Use the shortlist to choose where to start, then check each scorecard for context, throughput, listed price, and benchmark details.

Top Overall Model
Gemini 4 Carbon 72.8
Top Coding Score
Claude Mythos 5.2 Coding 93.0
Top Open Weights
Kimi K3.1 Pro Overall 63.2
Fastest Listed Output
Gemini 3.5 Flash 376 tps

A practical starting point

Shortlist by workload

Each score row shows its two highest-scoring models; the open-weight row ranks overall scores within that group. Scores help narrow the field, but do not predict every prompt or production setup.

Software engineering

Shortlist models with stronger coding results before comparing tool support and cost.

Top Coding scoreClaude Mythos 5.293.0
Next: Claude Fable 5.2 92.4Gap: 0.6 pts

Reasoning and math

Start with the reasoning pillar, then inspect the individual evaluations that matter to your work.

Top Reasoning scoreGemini 4 Carbon97.5
Next: Claude Mythos 5.2 96.5Gap: 1.0 pts

Agent and tool workflows

Use the agent score as a first pass; tool reliability can vary by task and setup.

Top Agent scoreGemini 4 Carbon95.4
Next: Claude Mythos 5.2 94.8Gap: 0.6 pts

Run with open weights

Ranked by overall score among open-weight models. Hosting and hardware costs are separate.

Top Overall score among open weightsKimi K3.1 Pro63.2
Next: Inkling-2-Small 61.8Gap: 1.4 pts

RankLLMs scores combine benchmark signals; speed and price are separate product specs. Read the scoring methodology.

Providers & Models Directory

Showing all 104 models across 24 providers

Google

11 models
Proprietary models

Anthropic

13 models
Proprietary models
#2 SILVER

Claude Mythos 5.2

→
70.8 ptsNEWProprietary weights
Context1M
Speed80 tps
Price / 1M tokens.00
Nov 12, 2026Scorecard Specs →
#3 BRONZE

Claude Fable 5.2

→
69.5 ptsNEWProprietary weights
Context1M
Speed85 tps
Price / 1M tokens.00
Nov 12, 2026Scorecard Specs →
#4

Claude Opus 5.5

→
68.6 ptsNEWProprietary weights
Context1M
Speed98 tps
Price / 1M tokens$4.00
Sep 22, 2026Scorecard Specs →
#7

Claude Fable 5.1

→
63.8 ptsNEWProprietary weights
Context1M
Speed68 tps
Price / 1M tokens$10.00
Sep 4, 2026Scorecard Specs →
#11

Claude Fable 5

→
62.0 ptsProprietary weights
Context1M
Speed110 tps
Price / 1M tokens$10.00
Jun 7, 2026Scorecard Specs →
#19

Claude Mythos Preview

→
58.2 ptsUNRELEASEDProprietary weights
Context1M
Speed50 tps
Price / 1M tokensFree/Local
Jan 15, 2025Scorecard Specs →
#22

Claude Opus 5

→
57.5 ptsProprietary weights
Context1M
Speed68 tps
Price / 1M tokens$5.00
Jul 24, 2026Scorecard Specs →
#36

Claude Opus 4.8

→
54.8 ptsProprietary weights
Context1M
Speed57 tps
Price / 1M tokens$5.00
May 28, 2026Scorecard Specs →
#37

Claude Sonnet 5

→
54.8 ptsProprietary weights
Context1M
Speed77 tps
Price / 1M tokens$2.00
Jun 29, 2026Scorecard Specs →
#41

Claude Opus 4.6

→
54.0 ptsProprietary weights
Context1M
Speed35 tps
Price / 1M tokens$5.00
Feb 4, 2026Scorecard Specs →
#54

Claude Sonnet 4.6

→
49.8 ptsProprietary weights
Context1M
Speed82 tps
Price / 1M tokens$3.00
Feb 17, 2026Scorecard Specs →
#66

Claude Opus 4.7

→
44.8 ptsProprietary weights
Context1M
Speed125 tps
Price / 1M tokens$5.00
Apr 14, 2026Scorecard Specs →
#86

Claude Opus 4.5

→
38.5 ptsProprietary weights
Context200K
Speed50 tps
Price / 1M tokens$5.00
Nov 24, 2025Scorecard Specs →

OpenAI

18 models
Proprietary models
#5

GPT-6 Astra

→
68.0 ptsNEWProprietary weights
Context1.1M
Speed73 tps
Price / 1M tokens$10.00
Sep 4, 2026Scorecard Specs →
#8

GPT-Rosalind-6

→
63.5 ptsNEWProprietary weights
Context1M
Speed72 tps
Price / 1M tokens.00
Oct 25, 2026Scorecard Specs →
#10

GPT-6 Cyber

→
62.5 ptsNEWProprietary weights
Context1.05M
Speed65 tps
Price / 1M tokens.00
Oct 28, 2026Scorecard Specs →
#16

GPT-6 Sol

→
60.5 ptsNEWProprietary weights
Context1.1M
Speed126 tps
Price / 1M tokens$2.00
Sep 22, 2026Scorecard Specs →
#25

GPT-5.6 Sol

→
57.2 ptsProprietary weights
Context1.1M
Speed89 tps
Price / 1M tokens$4.00
Jul 9, 2026Scorecard Specs →
#35

GPT-5.6 Terra

→
55.0 ptsProprietary weights
Context1.1M
Speed117 tps
Price / 1M tokens$2.00
Jul 9, 2026Scorecard Specs →
#42

GPT-5.4

→
53.8 ptsProprietary weights
Context1.1M
Speed14 tps
Price / 1M tokens$2.50
Mar 5, 2026Scorecard Specs →
#48

GPT-5.5

→
52.2 ptsProprietary weights
Context1.1M
Speed33 tps
Price / 1M tokens$5.00
Apr 23, 2026Scorecard Specs →
#51

GPT-6 Luna

→
51.2 ptsNEWProprietary weights
Context1.1M
Speed157 tps
Price / 1M tokens$0.10
Sep 22, 2026Scorecard Specs →
#55

GPT-5.6 Luna

→
49.8 ptsProprietary weights
Context1.1M
Speed157 tps
Price / 1M tokens$0.20
Jul 9, 2026Scorecard Specs →
#64

GPT-5.5 Pro

→
45.5 ptsProprietary weights
Context1.1M
Speed85 tps
Price / 1M tokens$30.00
Apr 23, 2026Scorecard Specs →
#70

GPT-5.2 Pro

→
43.4 ptsProprietary weights
Context400K
Speed95 tps
Price / 1M tokens$21.00
Dec 11, 2025Scorecard Specs →
#74

GPT-5.2

→
42.0 ptsProprietary weights
Context400K
Speed50 tps
Price / 1M tokens$1.75
Dec 11, 2025Scorecard Specs →
#91

GPT-5.3 Codex

→
37.1 ptsProprietary weights
Context400K
Speed18 tps
Price / 1M tokens$1.75
Feb 5, 2026Scorecard Specs →
#93

GPT-5.1 Thinking

→
36.9 ptsProprietary weights
Context400K
Speed50 tps
Price / 1M tokens$1.25
Nov 12, 2025Scorecard Specs →
#94

GPT-5.1 Instant

→
36.8 ptsProprietary weights
Context128K
Speed50 tps
Price / 1M tokens$1.10
Nov 12, 2025Scorecard Specs →
#95

GPT-5.1

→
36.7 ptsProprietary weights
Context400K
Speed270 tps
Price / 1M tokens$1.25
Nov 13, 2025Scorecard Specs →
#104

GPT-5.1 High

→
35.7 ptsProprietary weights
Context400K
Speed120 tps
Price / 1M tokens$2.22
Jan 15, 2025Scorecard Specs →

Meta

7 models
Proprietary models

Moonshot AI

6 models
Open-weight models

Thinking Machines Lab

2 models
Open-weight models

Alibaba Qwen

1 model
Proprietary models

DeepSeek

7 models
Open-weight models

MiniMax

4 models
Open-weight models

StepFun

2 models
Open-weight & proprietary

InclusionAI

1 model
Open-weight models

Xiaomi

4 models
Open-weight & proprietary

Mistral AI

1 model
Open-weight models

Zhipu AI

5 models
Open-weight models

Qwen / Alibaba

8 models
Open-weight & proprietary

xAI

3 models
Proprietary models

Tencent

2 models
Open-weight models

Sarvam AI

1 model
Open-weight models

ByteDance

3 models
Proprietary models

Sakana AI

1 model
Proprietary models

Poolside

1 model
Open-weight models

NVIDIA

1 model
Open-weight models

Upstage AI

1 model
Proprietary models

Meituan

1 model
Open-weight models