Model · checked 2026-08-21
At a glance
What it's for
Best for: Self-hosted general assistant, RAG, and internal tools.
Not ideal for: Frontier reasoning or multimodal tasks.
Role: Open-weight generalist
Ways to use this model
Each row is a separate provider route, plan, or model download. A dash means the fact is unknown, never free.
Swipe sideways to see price, availability, build, and comparison actions.
| Plan or version | Route or download | Provider | Delivery | Source identity | Price coverage | Availability | Verified | Build / compare |
|---|---|---|---|---|---|---|---|---|
| Llama 3.3 70B - GGUF Q4_K_M | Llama 3.3 70B GGUF Q4_K_M through LM Studio | LM Studio | Model download | lmstudio-community/Llama-3.3-70B-Instruct-GGUF:Q4_K_M | — | Not observed | 2026-08-26 | Checking fit…Compare |
| Llama 3.3 70B | Llama 3.3 70B through OpenRouter | OpenRouter | Hosted API | meta-llama/llama-3.3-70b-instruct | $0.10 / $0.32 per M | available | 2026-08-21 | Checking fit…Compare |
Specifications
Swipe sideways to see every specification, source, and verification date.
| Specification | Value | Source | Verified |
|---|---|---|---|
| API format | OpenAI | OpenRouter | 2026-07-06 |
| Context window | 131072 | OpenRouter | 2026-07-06 |
| Total parameters | 70B | Meta official Llama 3.3 model card | 2026-08-22 |
| License | Llama 3.3 Community License Agreement | Meta official Llama 3.3 model card | 2026-08-22 |
| Restricted access | No | OpenRouter | 2026-07-06 |
| Capabilities | Tools | OpenRouter | 2026-07-06 |
| Execution Paths | api, self_host | OpenRouter | 2026-07-06 |
Price evidence
| Type | USD | Basis | Conditions | Source | Verified | Freshness |
|---|---|---|---|---|---|---|
| per_million_in | $0.10 | required · reported | OpenRouter default price-weighted routing observation; this pair matched the AkashML FP8 Llama 3.3 70B endpoint on 2026-07-29, but the actual routed provider and billed rate can vary with availability, quantization, and routing preferences. | src_model_price_api | 2026-08-21 | vendor · 31d |
| per_million_out | $0.32 | required · reported | OpenRouter default price-weighted routing observation; this pair matched the AkashML FP8 Llama 3.3 70B endpoint on 2026-07-29, but the actual routed provider and billed rate can vary with availability, quantization, and routing preferences. | src_model_price_api | 2026-08-21 | vendor · 31d |
Cost at three usage tiers
| A few questions and small tasks each day. | $0.89/mo |
| Daily use, plus some scheduled tasks. | $2.96/mo |
| Working for you most of the day, including web tasks. | $8.88/mo |
Token cost only; add a host for the full-stack estimate.