Cheapest leaderboard
Live ranking of providers per model. Filter by region, context window, latency budget.
View ranking →Pick your GPU or enter your VRAM, and modelbeat ranks the open-weight LLMs that fit — at the best quantization for your card, with a composite intelligence index, output speed, and context window. Can't run it locally? We show the cheapest provider to rent it instead. 50 models across 7 providers, scraped nightly.
Live ranking of providers per model. Filter by region, context window, latency budget.
View ranking →Every provider stacked, 90-day price history, observed latency, benchmarks side-by-side.
Browse models →Compare hosts on price, capability, throughput and a workload of your choice.
Open providers →Plug in monthly tokens — get the actual bill on every provider. No marketing fluff.
Run a workload →| Model | Provider | Before | After | Δ | Tier | Trend |
|---|---|---|---|---|---|---|
| DeepSeek V3.2 | DeepInfra | $1.10 | $0.99 | -10.0% | output | |
| Llama 4 70B | Hyperbolic | $0.42 | $0.40 | -4.8% | output | |
| Qwen 3 72B | Together | $0.95 | $0.90 | -5.3% | input | |
| Mistral Large 3 | Bedrock | $5.00 | $4.40 | -12.0% | output | |
| Gemma 3 27B | Groq | $0.30 | $0.27 | -10.0% | input |
Every Monday, the biggest price drops, new provider listings and notable benchmark shifts — one short email, unsubscribe with one click.