LLM Stats
Composite index published by LLM Stats. It is not a percentage of correct answers.
Compare published indexes, API pricing and context limits. Pick two to four options to explore how they differ.
Snapshot: 2026-09-08 · Public sources · Not a real-time feedAtlas · version 1 · Reviewed on 2026-09-08 · Not a real-time feed · Evidence d15f1a66e958
| Metric | DeepSeek V4 Pro | Claude Fable 5.1 |
|---|---|---|
| LLM Statspoints · higher is better | No dataNot included | 56.8LLM Stats ↗ |
| Reasoningpoints · higher is better | No dataNot included | 53.9LLM Stats ↗ |
| Codingpoints · higher is better | No dataNot included | 36LLM Stats ↗ |
| Input priceUSD / M tokens · lower is better | $1.32Vendor ↗ | $10.00Vendor ↗ |
| Output priceUSD / M tokens · lower is better | $3.96Vendor ↗ | $50.00Vendor ↗ |
| Contexttokens · higher is better | 1.00MVendor ↗ | 1.00MVendor ↗ |
| Max outputtokens · higher is better | 384KVendor ↗ | 128KVendor ↗ |
Standard API, USD per million tokens, no caching, short context. Indexes are an LLM Stats snapshot; they are not percentages or Codifly tests. A curated catalog, not an exhaustive one.
Composite index published by LLM Stats. It is not a percentage of correct answers.
A relationship between two dimensions, not an automatic recommendation.
Input × rate + output × rate. Change the volume to see the difference.
Aggregate volume of short-context requests. No caching, tools, taxes, batch or long-context surcharges. Not a quote.
| Feature | DeepSeek V4 Pro | Claude Fable 5.1 |
|---|---|---|
| Contexto publicado | 1M; salida máxima publicada: 384K | No data |
| Entrada | Texto | No data |
| Identificador API | deepseek-v4-pro | No data |
| Tarifa de referencia | Pico, cache miss; fuera de pico 50% | No data |
| Versión consultada | DeepSeek-V4-Pro-0813 | No data |
| Distribución | No data | API propietaria |
| Entradas | No data | Texto, imagen |
| Límite de contexto | No data | Ventana publicada: 1.000.000 tokens |
| Salidas | No data | Texto |
What to validate before you make a decision.
Se muestra tarifa PICO sin caché. Horas pico: lunes a viernes 01–04 y 06–10 UTC. Fuera de pico cuesta la mitad. Las cifras 1M/384K conservan la notación publicada; los tokens de razonamiento consumen presupuesto.
Checked on 2026-09-08 · Editorial updateTarifa base de la API directa. Caché, batch, herramientas y otros proveedores tienen condiciones propias. El índice no determina el resultado en tu caso de uso.
Checked on 2026-09-08 · Editorial updateShare your requirements and the options you're weighing. No sign-up, no automatic subscriptions.