Open multimodal Llama for strong reasoning with efficient everyday serving
Llama 4 Maverick 17B Instruct by meta is offered by 8 providers on this page. Its reference price is $0.53 per 1M input tokens and $1.62 per 1M output tokens. The lowest paid channel is Abacus at $0.14 / $0.59 per 1M, about 3.8× below the reference price. That channel is a third-party gateway, so confirm its availability and rate limits before depending on it.
The context window is 131,072 tokens at the reference host, but hosts report different limits, from 8,192 to 1,048,576, so the usable window depends on the provider you pick. It supports tool use. Accepted input modalities are Text and Image. The weights are open, so it can also be self-hosted or served through a gateway of your choice. Providers report the capability flags inconsistently, so verify a specific feature against the host you plan to use. Its training knowledge cuts off at 2024-08.
| Provider | Tier | Input | Output | Cache read | Cache write | Context | Output limit | Status |
|---|---|---|---|---|---|---|---|---|
| Abacus meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8 | Gateway | $0.14 | $0.59 | — | — | 1,048,576 ⚠ | 8,192 | |
| DevPass (LLM Gateway) | Gateway | $0.27 | $0.85 | — | — | 1,048,576 ⚠ | 2,048 | |
| LLM Gateway | Gateway | $0.27 | $0.85 | — | — | 1,048,576 ⚠ | 8,192 | |
| Amazon Bedrock meta.llama4-maverick-17b-instruct-v1:0 | Cloud | $0.24 | $0.97 | — | — | 1,000,000 ⚠ | 16,384 | |
| LLM Gateway | Gateway | $0.24 | $0.97 | — | — | 8,192 ⚠ | 2,048 | |
| Charm Hyper llama-4-maverick-17b-128e-instruct-fp8 | Gateway | $0.274 | $0.899 | — | $0.137 | 430,000 ⚠ | 43,000 | |
| Neon llama-4-maverick | Cloud | $0.50 | $1.50 | — | — | 1,000,000 ⚠ | 8,192 | |
| LLM Gateway | Gateway | $0.53 | $1.62 | — | — | 131,072 | 8,192 |
Sorted by blended price (input×0.75 + output×0.25) asc. The official channel always shows regardless of rank. Whether a gateway's low price is actually usable can't be verified.
| Name | Conditions | Score | Metric | Source |
|---|---|---|---|---|
| Aider Polyglot | — | 15.6 | percent correct | Source ↗ |
| SWE-Bench Pro | dataset: public | 5.24 | resolve rate | Source ↗ |
The same benchmark scores very differently across harness / dataset, so the qualifying conditions must be shown together.