AI SEARCH CITATION ATLAS · OBJECT 06
Current AI model price and capability data, with the scope left intact.
This snapshot lists 417 OpenRouter models, their published prices and API features, plus the API's current top-weekly order.
asca:openrouter-model-atlas@2026-08-27.v1Popularity, price, and compatibility are separate signals.
OpenRouter's top-weekly order reflects its own recent token processing. Price and supported parameters come from the same catalog response, but neither measures global demand, provider-direct availability, or task quality.
OPENROUTER TOP-WEEKLY ORDER
Current order with listed price and API features
| Rank | Model | Input / 1M | Output / 1M | Context | Tools | Structured |
|---|---|---|---|---|---|---|
| 1 | DeepSeek: DeepSeek V4 Flash 0731 | $0.06 | $0.12 | 1.31M | Yes | Yes |
| 2 | Xiaomi: MiMo-V2.5 | $0.14 | $0.28 | 1.05M | Yes | Yes |
| 3 | Tencent: Hy3 | $0.132 | $0.528 | 262.14K | Yes | Yes |
| 4 | DeepSeek: DeepSeek V4 Flash 0423 | $0.0795 | $0.159 | 1.05M | Yes | Yes |
| 5 | NVIDIA: Nemotron 3 Ultra (free) | $0 | $0 | 1M | Yes | No |
Price conversion: published USD per token × 1,000,000. “Top weekly” is the API response order, not an exact public token-total table.
LARGEST LISTED CONTEXT WINDOWS
Five priced entries with the most context
| Top-weekly rank | Model | Input / 1M | Output / 1M | Context | Tools | Structured |
|---|---|---|---|---|---|---|
| 102 | SpaceXAI: Grok 4.20 | $1.25 | $2.50 | 2M | Yes | Yes |
| 187 | SpaceXAI: Grok 4.20 Multi-Agent | $1.25 | $2.50 | 2M | No | Yes |
| 388 | DeepSeek V4 Flash Latest | $0.03 | $0.075 | 1.31M | Yes | Yes |
| 1 | DeepSeek: DeepSeek V4 Flash 0731 | $0.06 | $0.12 | 1.31M | Yes | Yes |
| 123 | Meta: Llama 4 Scout | $0.11 | $0.34 | 1.31M | Yes | Yes |
Router entries without a listed non-negative input and output price are excluded from this table.
One public API response. No inferred market totals.
- Fetch OpenRouter's public Models API with
sort=top-weekly. - Publish the fetched JSON unchanged inside a dated source snapshot.
- Copy the model ID, catalog rank, listed prices, context limit, and declared supported parameters into CSV.
- Convert only the displayed price unit from per token to per million tokens.
What these numbers can and cannot establish.
- Observed range: 417 models returned by OpenRouter at 2026-08-27T00:23:25.942Z; the weekly order covers OpenRouter processing only.
- Bias: models and usage outside OpenRouter are excluded, while models can use different upstream tokenizers and routing configurations.
- Limit: the public endpoint provides weekly order, not each model's precise weekly token total; catalog metadata is not a quality benchmark.
- Confidence: high for rerunning this recorded response and conversion; Limited for claims beyond this catalog snapshot.
DOWNLOAD · VERIFY · CITE
Carry the timestamp with the model claim.
6fa50aede90a6d0cfebe04d18c650728fe119bb58800b561b85414840f68c03f“This page reports an OpenRouter catalog snapshot, not a global AI model ranking.”— asca:openrouter-model-atlas@2026-08-27.v1
Ten practical views of the same snapshot.
- Cheapest paid LLM API models
- Long-context LLMs for a 128K + 4K workload
- Tool-calling compatible models
- Structured-output compatible models
- Vision-language models
- Free LLM API models
- Most popular AI models by OpenRouter weekly order
- Fastest LLM API models by OpenRouter order
- Coding AI models by OpenRouter order
- Agentic AI models by OpenRouter order
AI search source overlap remains a separate measurement.
Read the Citation Atlas source-overlap analysis. It measures cited-domain overlap in a May 2026 snapshot; it does not measure model catalog prices, API compatibility, or OpenRouter usage.