Best overall
Closed model
Anthropic
- Intelligence
- 59.9
- Speed
- 70tok/s
- Input / output
- $10 / $50
- Context
- 1M
Best fit
High-stakes reasoning and complex agentic workflows
AI · LEADERBOARD
Choisissez le modèle adapté au travail.
Une comparaison pratique des modèles IA réellement déployables aujourd’hui.
Snapshot: 21 Jul 2026 · 10:01 TST12 models compared
Independent editorial snapshot · not live platform telemetry
The short answer
Three leaders, three different decisions.
Best open model
Open weights
Z AI
- Intelligence
- 51.1
- Speed
- 198tok/s
- Input / output
- $1.4 / $4.4
- Context
- 1M
Best fit
Sovereign deployments that still need strong reasoning
Best value
Open weights
Xiaomi
- Intelligence
- 37.2
- Speed
- 64tok/s
- Input / output
- $0.14 / $0.29
- Context
- 1.05M
Best fit
Cost-sensitive extraction and high-volume structured work
OpenRouter usage
OpenRouter LLM usage ranking
Top models by prompt, completion, and reasoning tokens processed on OpenRouter. Showing the first 20 only.
Top 20OpenRouter data · cached hourlyUpstream · Jul 21, 2026
- 1.Hy3 (free)by Tencent10.2T tokens24%
- 2.MiMo-V2.5by Xiaomi9.43T tokens29%
- 3.DeepSeek V4 Flashby DeepSeek5.37T tokens2%
- 4.GLM 5.2by Z.ai3.63T tokens16%
- 5.MiniMax M3by MiniMax3.23T tokens22%
- 6.DeepSeek V4 Proby DeepSeek2.8T tokens9%
- 7.Nemotron 3 Ultra (free)by Nvidia2.67T tokens2%
- 8.Claude Opus 4.7by Anthropic1.93T tokens8%
- 9.Claude Opus 4.8by Anthropic1.91T tokens13%
- 10.Claude Sonnet 5by Anthropic1.13T tokens21%
- 11.Gemini 3 Flash Previewby Google963B tokens3%
- 12.Step 3.7 Flashby StepFun961B tokens5%
- 13.Claude Sonnet 4.6by Anthropic877B tokens8%
- 14.Kimi K3by Moonshot AI741B tokensnew
- 15.Hy3by Tencent638B tokens>999%
- 16.GPT-5.5 (batch)by OpenAI599B tokens22%
- 17.Gemini 2.5 Flashby Google591B tokens3%
- 18.Laguna M.1 (free)by Poolside575B tokens3%
- 19.MiMo-V2.5-Proby Xiaomi572B tokens12%
- 20.Gemini 2.5 Flash Liteby Google570B tokens2%
From benchmark to decision
Which model for which job?
Nine recurring production scenarios, with the trade-offs that matter before a team commits to a model.
Selected use case
01 / 09
Decision view
Customer support
Why it fits
- 01294 tok/s keeps wait time low
- 02Multilingual intake
- 03Strong multimodal routing
How the ranking works
Transparent enough to disagree with.
The ranking combines public benchmark signals with price, measured throughput, context and deployment constraints. It is a decision aid—not a universal definition of model quality.
- Dataset snapshot
- 21 Jul 2026 · 10:01 TST
- Review cadence
- Revalidated when major model data changes
Value score
(intelligence − 25)² × min(1, speed ÷ 30) ÷ (input cost + output cost × 3)
Output is weighted ×3 to reflect output-heavy production use; intelligence is squared so cheap but weak models do not dominate.
- Intelligence
- Public benchmark and reasoning signals form the default order.
- Speed
- Output tokens per second; provider and region still affect real latency.
- Cost
- Public input and output price per million tokens, before enterprise discounts.
- Deployment
- Context, weight access and hosting constraints determine production fit.
Public references
Prices and performance can change without notice. Validate the finalists against your own prompts, infrastructure and risk requirements before procurement.
FAQ
Questions worth asking before the benchmark wins the argument.
Claude Fable 5 leads the intelligence-first ranking in this captured dataset. The best choice can change once price, latency, hosting and task-specific evaluation enter the decision.
GLM-5.2 is the leading open-weight option in this snapshot, combining a strong intelligence signal, high throughput and sovereign deployment flexibility.
MiMo-V2.5 leads the value formula because its low token price and large context offset a lower raw intelligence score. It is particularly compelling for structured, high-volume work.
No. Context capacity only describes how much can be submitted. Retrieval design, instruction quality, attention behaviour and evaluation still determine whether that context produces a useful answer.
This page is an editorial snapshot, not live platform telemetry. We revalidate the dataset when major model releases or material pricing and performance changes occur.
The default order uses the public intelligence signal. Speed, price, context, openness and a separate value formula remain visible so teams can challenge the default order.
Tenten does not sell model tokens through this page. Public sources are linked, assumptions are shown, and we recommend a task-specific evaluation before any production decision.
TENTEN AI · MODEL EVALUATION
The leaderboard is the start
Bring the workflow. We’ll test the shortlist.
A model only becomes a good decision after it survives your documents, permissions, latency budget and failure cases.