Qwen3-Embedding-8B
Alibaba Qwen · open weights
$0.010 per 1M input tokens, cheapest of 3 sellers (DeepInfra)
What was measured
| Board | Mean (Task) | Mean (TaskType) | Tasks | Revision |
|---|---|---|---|---|
| MTEB(eng, v2) | 75.23 | 68.7 | 41 of 41 | 4e423935c619ae4df87b646a3ce949610c66241c |
By task type: Classification 90.43 · Clustering 58.57 · PairClassification 87.52 · Reranking 51.56 · Retrieval 69.44 · STS 88.58 · Summarization 34.83
What sellers charge
Spread across 3 distinct sellers: 10× between cheapest and dearest at the reference unit.
| Seller | As the seller states it | Conditions | Normalised per 1M input tokens | Source |
|---|---|---|---|---|
| DeepInfra | 1e-8 per_token | — | $0.010 per token × 1,000,000 | api.deepinfra.com read 2026-09-19 · feed |
| OpenRouter | 1e-8 per_token | context 32768 | $0.010 per token × 1,000,000 | openrouter.ai read 2026-09-19 · feed |
| Fireworks | 0.1 per_million_tokens | — | $0.100 as stated | fireworks.ai read 2026-09-06 · verified |
| Embeddings table, "$ / 1M input tokens": "Qwen3 8B | $0.1". The table's other two rows price by base-model parameter count (up to 150M $0.008; 150M-350M $0.016) and are not recorded, because assigning a model to a band would need a parameter count this page does not give. | ||||
Provenance
- Quality: https://github.com/embeddings-benchmark/results, licensed cc0-1.0.
- Prices: https://api.deepinfra.com/models/list, https://openrouter.ai/api/v1/embeddings/models, https://fireworks.ai/pricing
- Built 2026-09-19 from the catalogue snapshot in this repository.
MTEB results, https://github.com/embeddings-benchmark/results, released under CC0-1.0; MTEB(eng, v2) means computed from the per-task files as described in data/modalities/embeddings/benchmark.json.