Model Tracker
The models that matter right now, in one table — curated editorially, so it reflects what's actually worth using rather than everything that exists.
Prices are per 1M tokens. Last reviewed 2026-07-15— “verify” means check the provider's current pricing page before budgeting.
| Model | Provider | Context | Input | Output | Weights | Best for |
|---|---|---|---|---|---|---|
| Claude Fable 5 | Anthropic | 1M | $10.00 | $50.00 | closed | Hardest reasoning and long-horizon agentic work; always-on thinking |
| Claude Opus 4.8 | Anthropic | 1M | $5.00 | $25.00 | closed | Flagship coding and agentic work; long autonomous runs |
| Claude Sonnet 5 | Anthropic | 1M | $3.00 ($2 intro) | $15.00 ($10 intro) | closed | Near-Opus coding quality at Sonnet cost; the volume workhorse |
| Claude Haiku 4.5 | Anthropic | 200K | $1.00 | $5.00 | closed | Fast, cheap classification, extraction, and routing |
| GPT-5.x (latest) | OpenAI | 400K+ | verify | verify | closed | General-purpose flagship; broad ecosystem and tooling |
| o-series (reasoning) | OpenAI | 200K+ | verify | verify | closed | Hard math, science, and multi-step reasoning at higher latency |
| Gemini 3 Pro | 1M+ | verify | verify | closed | Very long context, multimodal breadth, Workspace integration | |
| Llama 4 family | Meta | up to 1M+ | self-host | self-host | open | Open-weights general models; fine-tuning base for enterprises |
| DeepSeek V3.x / R1 | DeepSeek | 128K | very low | very low | open | Open-weights reasoning at aggressive cost; MoE efficiency |
| Qwen 3 family | Alibaba | 128K+ | self-host | self-host | open | Strong multilingual open weights across many sizes; agent tuning |
| Mistral (Large / small) | Mistral | 128K | verify | verify | open | European provider; efficient open and hosted models |
| Gemma 3 family | 128K | self-host | self-host | open | Small open models that run on a single GPU or laptop |