Leaderboard / best model for classify / route
Best model for classify / route
For classify / route, the best value is Jev (TypeSafe) — 100% cheaper than Claude Fable 5.1 at near-equal quality.
Best value
Jev (TypeSafe)
Highest quality
Claude Fable 5.1
Cheapest
Jev (TypeSafe)
All models for classify / route, ranked by quality-per-dollar
| Model | Provider | Capability | Cost / call |
|---|---|---|---|
| Jev (TypeSafe)BEST VALUE | TypeSafe | 92 | $0.0001 |
| GLM 5.3 Flash | Z.ai | 88 | $0.0003 |
| Qwen 3.8 Flash | Alibaba | 88 | $0.0005 |
| Grok 4.1 Fast | xAI | 89 | $0.0006 |
| DeepSeek V4 Flash | DeepSeek | 89 | $0.0007 |
| Llama 4 Maverick | Meta | 87 | $0.0006 |
| GPT-5.6 Luna | OpenAI | 91 | $0.0009 |
| DeepSeek V4.1 Flash | DeepSeek | 90 | $0.0010 |
| GPT-5 mini | OpenAI | 90 | $0.0014 |
| Mistral Large 3 | Mistral | 89 | $0.0015 |
| Gemini 2.5 Flash | 89 | $0.0017 | |
| DeepSeek V4 Pro | DeepSeek | 91 | $0.0020 |
| Gemini 3.8 Flash | 91 | $0.0030 | |
| Gemini 3.7 Flash | 91 | $0.0030 | |
| Gemini 3.6 Flash | 90 | $0.0030 | |
| Qwen 3 Max | Alibaba | 88 | $0.0031 |
| Claude Haiku 4.5 | Anthropic | 90 | $0.0040 |
| Muse Spark 1.3 | Meta | 88 | $0.0040 |
| GLM 5.3 | Z.ai | 90 | $0.0043 |
| Grok 4.6 | xAI | 92 | $0.0060 |
| Qwen 3.8 Max | Alibaba | 89 | $0.0060 |
| GPT-5 | OpenAI | 94 | $0.0069 |
| Gemini 2.5 Pro | 91 | $0.0069 | |
| GPT-5.6 Sol | OpenAI | 95 | $0.0080 |
| Claude Sonnet 5 | Anthropic | 93 | $0.0080 |
| GPT-5.6 Terra | OpenAI | 94 | $0.0090 |
| Gemini 3.1 Pro | 93 | $0.0090 | |
| GPT-5.2 | OpenAI | 94 | $0.0096 |
| Claude Sonnet 4.6 | Anthropic | 92 | $0.0120 |
| Kimi K3 | Moonshot | 90 | $0.0120 |
| Claude Opus 5 | Anthropic | 94 | $0.0200 |
| Claude Opus 4.8 | Anthropic | 93 | $0.0200 |
| GPT-5.5 | OpenAI | 94 | $0.0225 |
| Claude Fable 5.1 | Anthropic | 95 | $0.0400 |
| GPT-6 Astra Pro | OpenAI | 95 | $0.0400 |
| GPT-6 Astra | OpenAI | 95 | $0.0400 |
| Claude Fable 5 | Anthropic | 94 | $0.0400 |
What is the best model for classify / route in 2026?
For classify / route, Jev (TypeSafe) offers the best quality-per-dollar — 100% cheaper than the premium default (Claude Fable 5.1) at near-equal capability. If cost is the only concern, Jev (TypeSafe) is the absolute floor.
How much can I save on classify / route?
At 10,000 calls/month, switching from Claude Fable 5.1 to Jev (TypeSafe) takes the cost from $400.00 to $0.6300 — about $399.37/month saved.
Route your classify / route automatically
Tokenokio picks the best-value model per task and runs it on your own key. Zero float.
Try the RouterIllustrative pricing · seed capability scores, refined from real usage.