Hard questions and reasoning
Difficult problems, judged by the Artificial Analysis Intelligence Index.
Models that have a measurement or a declared feature for this task, with the evidence and the cost of a standard job. The size of the job is our assumption.
Open in the catalog with this filter: 71 Compare the first models All free models
- Models listed
- 66
- Measure used for the order
- Intelligence Index
List entries: 71. A free route of a model is its own entry, next to the paid model.
Our assumed job. Requests: 10. Tokens per request: 3,000 in, 8,000 out.
Cost of this job. Our assumption for this job. Requests: 10. Input tokens per request: 3,000. Output tokens per request: 8,000. Share of input served from cache: 0%. The prices are real; the size of the job is not. Thinking is billed as output tokens; a model that thinks longer costs more than this estimate.
Rank among scored models. A band label appears only where at least 30 models are scored. Ranks are among all models we list, including older ones. The share is rounded up to whole models, so a rank slightly past the exact share still counts. Models with equal scores share the better rank.
No free model reaches the top 25% here.
Free route. Rate limits apply and the route can differ from the paid one.
Free Hard questions and reasoning: Nemotron 3.5 Super VL · Rank 50 of 66, among scored models.
Top 10% 7
-
Claude Fable 5.2
Artificial AnalysisIntelligence Index: 64.6 · Rank 1 of 66, among scored models
No cheaper model scores as high
$3.44 Cost of the job $16.00 · mixed / 1M tokens -
GPT-6.2 Astra
Artificial AnalysisIntelligence Index: 62.1 · Rank 2 of 66, among scored models
$3.44 Cost of the job $16.00 · mixed / 1M tokens Price rises above 272k tokens -
Gemini 4 Neon
Artificial AnalysisIntelligence Index: 61.4 · Rank 3 of 66, among scored models
No cheaper model scores as high
$1.29 Cost of the job $6.00 · mixed / 1M tokens -
Gemini 4 Carbon
Artificial AnalysisIntelligence Index: 60.0 · Rank 4 of 66, among scored models
$0.86 Cost of the job $4.00 · mixed / 1M tokens promo -
Claude Opus 5.5
Artificial AnalysisIntelligence Index: 59.1 · Rank 5 of 66, among scored models
Models that score at least as high for less in this scenario: 2
$1.72 Cost of the job $8.00 · mixed / 1M tokens -
GPT-6.2 Sol
Artificial AnalysisIntelligence Index: 57.3 · Rank 6 of 66, among scored models
$0.86 Cost of the job $4.00 · mixed / 1M tokens Price rises above 272k tokens -
GPT-6 Astra
Artificial AnalysisIntelligence Index: 56.9 · Rank 7 of 66, among scored models
Models that score at least as high for less in this scenario: 6
$4.30 Cost of the job $20.00 · mixed / 1M tokens Price rises above 272k tokens
Top 25% 10
-
GPT-6.1 Sol
Artificial AnalysisIntelligence Index: 54.6 · Rank 8 of 66, among scored models
$0.86 Cost of the job $4.00 · mixed / 1M tokens Price rises above 272k tokens -
Claude Sonnet 5.5
Artificial AnalysisIntelligence Index: 54.5 · Rank 9 of 66, among scored models
$0.86 Cost of the job $4.00 · mixed / 1M tokens -
Gemini 4 Hydrogen
Artificial AnalysisIntelligence Index: 53.9 · Rank 10 of 66, among scored models
No cheaper model scores as high
$0.65 Cost of the job $3.00 · mixed / 1M tokens -
GPT-6.1 Terra
Artificial AnalysisIntelligence Index: 51.3 · Rank 11 of 66, among scored models
No cheaper model scores as high
$0.43 Cost of the job $2.00 · mixed / 1M tokens Price rises above 272k tokens -
Gemini 4 Argon
Artificial AnalysisIntelligence Index: 50.4 · Rank 12 of 66, among scored models
Models that score at least as high for less in this scenario: 2
$0.86 Cost of the job $4.00 · mixed / 1M tokens promo -
Claude Fable 5.1
Artificial AnalysisIntelligence Index: 50.0 · Rank 13 of 66, among scored models
Models that score at least as high for less in this scenario: 11
$4.30 Cost of the job $20.00 · mixed / 1M tokens -
Kimi K3.1 Pro
Artificial AnalysisIntelligence Index: 46.5 · Rank 14 of 66, among scored models
Models that score at least as high for less in this scenario: 2
$0.86 Cost of the job $4.00 · mixed / 1M tokens -
GPT-6 Sol
Artificial AnalysisIntelligence Index: 44.2 · Rank 15 of 66, among scored models
Models that score at least as high for less in this scenario: 2
$0.86 Cost of the job $4.00 · mixed / 1M tokens Price rises above 272k tokens -
Claude Opus 5
Artificial AnalysisIntelligence Index: 44.2 · Rank 15 of 66, among scored models
Models that score at least as high for less in this scenario: 11
$2.15 Cost of the job $10.00 · mixed / 1M tokens -
Muse Wonder
Artificial AnalysisIntelligence Index: 43.6 · Rank 17 of 66, among scored models
Models that score at least as high for less in this scenario: 10
$1.72 Cost of the job $8.00 · mixed / 1M tokens
Scored, outside the top 25% 49
-
Qwen4 Max
Artificial AnalysisIntelligence Index: 42.9 · Rank 18 of 66, among scored models
$0.38 Cost of the job $2.14 · mixed / 1M tokens promo -
Muse Spark 1.3
Artificial AnalysisIntelligence Index: 36.6 · Rank 19 of 66, among scored models
$0.38 Cost of the job $2.00 · mixed / 1M tokens -
Claude Haiku 5.5
Artificial AnalysisIntelligence Index: 36.5 · Rank 20 of 66, among scored models
$0.04 Cost of the job $0.20 · mixed / 1M tokens Price rises above 100k tokens -
Muse Ember
Artificial AnalysisIntelligence Index: 36.0 · Rank 21 of 66, among scored models
$0.04 Cost of the job $0.20 · mixed / 1M tokens -
Grok 4.7
Artificial AnalysisIntelligence Index: 35.5 · Rank 22 of 66, among scored models
Models that score at least as high for less in this scenario: 5
$0.54 Cost of the job $3.00 · mixed / 1M tokens Price rises above 200k tokens -
Hy4
Artificial AnalysisIntelligence Index: 34.6 · Rank 23 of 66, among scored models
Models that score at least as high for less in this scenario: 2
$0.11 Cost of the job $0.6254 · mixed / 1M tokens promo -
Qwen3.8 Max (0902)
Artificial AnalysisIntelligence Index: 33.7 · Rank 24 of 66, among scored models
Models that score at least as high for less in this scenario: 6
$0.54 Cost of the job $3.00 · mixed / 1M tokens -
DeepSeek V4.1 Flash 1017
Artificial AnalysisIntelligence Index: 32.6 · Rank 25 of 66, among scored models
Models that score at least as high for less in this scenario: 2
$0.11 Cost of the job $0.525 · mixed / 1M tokens Cheaper off-peak -
GPT-5.6 Terra
Artificial AnalysisIntelligence Index: 32.3 · Rank 26 of 66, among scored models
Models that score at least as high for less in this scenario: 17
$1.02 Cost of the job $4.50 · mixed / 1M tokens Price rises above 272k tokens -
GLM 5.3
Artificial AnalysisIntelligence Index: 31.7 · Rank 27 of 66, among scored models
Models that score at least as high for less in this scenario: 6
$0.39 Cost of the job $2.15 · mixed / 1M tokens -
Grok 4.6
Artificial AnalysisIntelligence Index: 31.0 · Rank 28 of 66, among scored models
Models that score at least as high for less in this scenario: 8
$0.54 Cost of the job $3.00 · mixed / 1M tokens Price rises above 200k tokens -
Gemini 3.8 Flash
Artificial AnalysisIntelligence Index: 30.8 · Rank 29 of 66, among scored models
Models that score at least as high for less in this scenario: 4
$0.32 Cost of the job $1.50 · mixed / 1M tokens promo -
Inkling 2 Small
Artificial AnalysisIntelligence Index: 30.6 · Rank 30 of 66, among scored models
Models that score at least as high for less in this scenario: 2
$0.09 Cost of the job $0.475 · mixed / 1M tokens via Cloudflare -
MiMo-V2.6-Pro
Artificial AnalysisIntelligence Index: 29.9 · Rank 31 of 66, among scored models
Models that score at least as high for less in this scenario: 2
$0.08 Cost of the job $0.5438 · mixed / 1M tokens -
Ling 3.1 Flash
Artificial AnalysisIntelligence Index: 29.6 · Rank 32 of 66, among scored models
$0.02 Cost of the job $0.09 · mixed / 1M tokens via Novita promo -
MiniMax M3.1 Flash
Artificial AnalysisIntelligence Index: 29.1 · Rank 33 of 66, among scored models
Models that score at least as high for less in this scenario: 3
$0.05 Cost of the job $0.2625 · mixed / 1M tokens Cheaper off-peak -
Qwen3.8 Flash
Artificial AnalysisIntelligence Index: 28.6 · Rank 34 of 66, among scored models
Models that score at least as high for less in this scenario: 1
$0.04 Cost of the job $0.23 · mixed / 1M tokens -
Step 5 Preview
Artificial AnalysisIntelligence Index: 28.5 · Rank 35 of 66, among scored models
Models that score at least as high for less in this scenario: 9
$0.25 Cost of the job $1.43 · mixed / 1M tokens -
GPT-6 Luna
Artificial AnalysisIntelligence Index: 27.9 · Rank 36 of 66, among scored models
Models that score at least as high for less in this scenario: 2
$0.04 Cost of the job $0.20 · mixed / 1M tokens Price rises above 272k tokens -
GLM 5.3 Flash
Artificial AnalysisIntelligence Index: 27.8 · Rank 37 of 66, among scored models
Models that score at least as high for less in this scenario: 5
$0.04 Cost of the job $0.2375 · mixed / 1M tokens -
MiMo-V2.6-Flash
Artificial AnalysisIntelligence Index: 26.7 · Rank 38 of 66, among scored models
Models that score at least as high for less in this scenario: 1
$0.03 Cost of the job $0.175 · mixed / 1M tokens -
Mistral Large 4
Artificial AnalysisIntelligence Index: 26.1 · Rank 39 of 66, among scored models
Models that score at least as high for less in this scenario: 12
$0.19 Cost of the job $1.05 · mixed / 1M tokens -
Muse Spark Open
Artificial AnalysisIntelligence Index: 24.9 · Rank 40 of 66, among scored models
Models that score at least as high for less in this scenario: 15
$0.38 Cost of the job $2.00 · mixed / 1M tokens via Fireworks -
GPT-5.6 Luna
Artificial AnalysisIntelligence Index: 23.8 · Rank 41 of 66, among scored models
Models that score at least as high for less in this scenario: 10
$0.10 Cost of the job $0.45 · mixed / 1M tokens Price rises above 272k tokens -
DeepSeek V4 Pro 0813
Artificial AnalysisIntelligence Index: 23.1 · Rank 42 of 66, among scored models
$1.98 mixed / 1M tokens in $1.32 · out $3.96 Cheaper off-peak Price disputed -
DeepSeek V4 Pro 0423
Artificial AnalysisIntelligence Index: 20.8 · Rank 43 of 66, among scored models
Models that score at least as high for less in this scenario: 13
$0.16 Cost of the job $1.10 · mixed / 1M tokens via SiliconFlow -
Inkling 2 Mini
Artificial AnalysisIntelligence Index: 20.1 · Rank 44 of 66, among scored models
Models that score at least as high for less in this scenario: 2
$0.03 Cost of the job $0.1588 · mixed / 1M tokens via PrimeIntellect -
Qwen3.8 27B
Artificial AnalysisIntelligence Index: 20.1 · Rank 44 of 66, among scored models
Models that score at least as high for less in this scenario: 16
$0.22 Cost of the job $0.9563 · mixed / 1M tokens -
MiniMax M3
Artificial AnalysisIntelligence Index: 18.9 · Rank 46 of 66, among scored models
Models that score at least as high for less in this scenario: 12
$0.11 Cost of the job $0.525 · mixed / 1M tokens -
Beam 1.0 501B A23B
Artificial AnalysisIntelligence Index: 17.7 · Rank 47 of 66, among scored models
Models that score at least as high for less in this scenario: 14
$0.11 Cost of the job $0.60 · mixed / 1M tokens via DigitalOcean -
Kimi K3
Artificial AnalysisIntelligence Index: 17.4 · Rank 48 of 66, among scored models
Models that score at least as high for less in this scenario: 38
$1.29 Cost of the job $6.00 · mixed / 1M tokens -
Solar Pro 4
Artificial AnalysisIntelligence Index: 14.2 · Rank 49 of 66, among scored models
Models that score at least as high for less in this scenario: 8
$0.05 Cost of the job $0.2625 · mixed / 1M tokens promo -
Nemotron 3.5 Super VL
Artificial AnalysisIntelligence Index: 13.6 · Rank 50 of 66, among scored models
Models that score at least as high for less in this scenario: 3
The free route differs in: Max output: 131k (other version: 944k).
$0.04 Cost of the job $0.175 · mixed / 1M tokens via Crusoe -
Inkling Small
Artificial AnalysisIntelligence Index: 12.9 · Rank 51 of 66, among scored models
Models that score at least as high for less in this scenario: 17
$0.11 Cost of the job $0.6375 · mixed / 1M tokens via DeepInfra -
Inkling
Artificial AnalysisIntelligence Index: 12.5 · Rank 52 of 66, among scored models
Models that score at least as high for less in this scenario: 24
$0.35 Cost of the job $1.76 · mixed / 1M tokens via Together -
Nemotron 3 Ultra
Artificial AnalysisIntelligence Index: 11.8 · Rank 53 of 66, among scored models
Models that score at least as high for less in this scenario: 21
The free route differs in: Context window: 1M (other version: 262k); Max output: 66k (other version: 16k); Structured output: Not listed (other version: Yes).
$0.19 Cost of the job $0.925 · mixed / 1M tokens via KT -
Gemini 3.5 Flash Lite
Artificial AnalysisIntelligence Index: 11.7 · Rank 54 of 66, among scored models
Models that score at least as high for less in this scenario: 22
$0.21 Cost of the job $0.85 · mixed / 1M tokens -
Claude Haiku 4.5
Artificial AnalysisIntelligence Index: 7.4 · Rank 55 of 66, among scored models
Models that score at least as high for less in this scenario: 31
$0.43 Cost of the job $2.00 · mixed / 1M tokens -
Muse Glimmer 30B
Artificial AnalysisIntelligence Index: 6.3 · Rank 56 of 66, among scored models
Models that score at least as high for less in this scenario: 14
$0.11 Cost of the job $0.525 · mixed / 1M tokens via DeepInfra -
Gemini 3.1 Flash Lite Preview
Artificial AnalysisIntelligence Index: 6.2 · Rank 57 of 66, among scored models
Models that score at least as high for less in this scenario: 20
$0.13 Cost of the job $0.5625 · mixed / 1M tokens -
Command A+
Artificial AnalysisIntelligence Index: 6.2 · Rank 57 of 66, among scored models
Models that score at least as high for less in this scenario: 21
$0.13 Cost of the job $0.60 · mixed / 1M tokens -
LFM3-1.7B
Artificial AnalysisIntelligence Index: 6.0 · Rank 59 of 66, among scored models
Models that score at least as high for less in this scenario: 1
$0.02 Cost of the job $0.11 · mixed / 1M tokens -
Nemotron 3.5 Lightning
Artificial AnalysisIntelligence Index: 5.5 · Rank 60 of 66, among scored models
The free route differs in: Context window: 1M (other version: 262k); Max output: 66k (other version: 236k); Structured output: Not listed (other version: Yes).
$0.01 Cost of the job $0.085 · mixed / 1M tokens via DeepInfra -
Granite 4.2 8B
Artificial AnalysisIntelligence Index: 5.1 · Rank 61 of 66, among scored models
$0.01 Cost of the job $0.0825 · mixed / 1M tokens via CoreWeave -
Mistral Medium 3.5
Artificial AnalysisIntelligence Index: 4.6 · Rank 62 of 66, among scored models
Ends January 31, 2027
Models that score at least as high for less in this scenario: 42
$0.65 Cost of the job $3.00 · mixed / 1M tokens -
Gemma 4 31B
Artificial AnalysisIntelligence Index: 4.4 · Rank 63 of 66, among scored models
Models that score at least as high for less in this scenario: 5
The free route differs in: Max output: 33k (other version: 16k); Structured output: Not listed (other version: Yes).
$0.03 Cost of the job $0.16 · mixed / 1M tokens via KT -
Mistral Small 4
Artificial AnalysisIntelligence Index: 4.3 · Rank 64 of 66, among scored models
Models that score at least as high for less in this scenario: 13
$0.05 Cost of the job $0.2625 · mixed / 1M tokens -
gpt-oss-120b
Artificial AnalysisIntelligence Index: 3.8 · Rank 65 of 66, among scored models
Models that score at least as high for less in this scenario: 2
$0.01 Cost of the job $0.0703 · mixed / 1M tokens via DeepInfra -
North Mini Code
Artificial AnalysisIntelligence Index: 3.4 · Rank 66 of 66, among scored models
Free
Method and sources
- How tasks are assigned
- A task chip is shown only when a third party has measured the model and it ranks in the top 10% or the top 25% of the scored models. Tasks matched by a declared feature (context size, input and output types) are listed on the task page and the model page, without a chip. We do not score models ourselves. The share is rounded up to whole models, so a rank slightly past the exact share still counts. Models with equal scores share the better rank.
- What "no cheaper model scores as high" means
- For a task we take one third-party score and the cost of a standard job (our assumption, printed beside every cost). A model is marked when no other model scores at least as high for less. It is a computed fact, not a recommendation.
- Artificial Analysis / Design Arena
- Intelligence, Coding and Agentic indices: Artificial Analysis (artificialanalysis.ai). Design categories: Design Arena (designarena.ai). Both are relayed by the OpenRouter public model list and were read on November 21, 2026. We compute ranks, tiers and costs from these numbers; the sources have not reviewed them. Scores belong to the model, not to the route it is served on. The sources do not publish how many comparisons or runs stand behind each number; new models can still move.
Artificial Analysis · Design Arena · Relayed by OpenRouter. opens in a new tab
- Fixed numbers we chose
- Top 10% and Top 25% bands, at least 30 scored models for a band label, long documents from 400k tokens of context, budget up to $0.30 mixed price per 1M tokens, "ending soon" within 30 days.