Qianfan-VL-8B
- Open weights
Prices
1 offering · prices as listedQianfan-VL-8B| Conditions/ 1K tokens | Input | Output |
|---|---|---|
| Standard past price, from the vendor’s page archived Jun 22, 2026 | ¥0.002 | ¥0.006 |
| Batch past price, from the vendor’s page archived Jun 22, 2026 | ¥0.0008 | ¥0.0024 |
Prices from 百度智能云千帆：模型价格 (opens in a new tab) (checked Oct 5, 2026)
Standard: the price a request with no conditions pays (the headline price elsewhere on the site).
Details
- Family
- qianfan-vl
- Model ID
Not stated on Baidu’s pages: max input, input modalities, output modalities, reasoning, reasoning efforts, reasoning can be off, max thinking budget, tool use, structured output, built-in tools, knowledge cutoff, licence, aliases.
Run locally
Weights: baidu/Qianfan-VL-8B- Parameters 8.8B
- Dense
- Native precision BF16
- Layers 32
- KV cache per token 128 KB
- Config positions 32K (config.json max_position_embeddings)
- Needs
- ≈ 18.2 GB (weights 16.4 (maker’s file) + KV cache 1.0 + overhead 0.8)
- Usable
- 21.6 GB (90% of its memory)
- Generation speed
- 27–40 tok/s
- Longest context
- 32K (the model’s limit)
- First token
- about 0.28–0.43 s for a 2K prompt
| Precision | Weights | Needs at 8K | On your hardware | Longest context |
|---|---|---|---|---|
| BF16 (native)maker’s file | 16.4 GB | ≈ 18.2 | ●Runs well 27–40 tok/s | 32K |
| Q8_0 | 8.7 GB | ≈ 10.2 | ●Runs well 48–72 tok/s | 32K |
| FP8 | 8.2 GB | ≈ 9.7 | ●Runs well 51–77 tok/s | 32K |
| Q6_K | 6.7 GB | ≈ 8.2 | ●Runs well 61–91 tok/s | 32K |
| Q5_K_M | 5.8 GB | ≈ 7.3 | ●Runs well 69–103 tok/s | 32K |
| Q5_0 | 5.6 GB | ≈ 7.1 | ●Runs well 71–106 tok/s | 32K |
| Q4_K_M | 5.0 GB | ≈ 6.5 | ●Runs well 78–118 tok/s | 32K |
| Q4_0 | 4.6 GB | ≈ 6.1 | ●Runs well 84–125 tok/s | 32K |
| IQ4_XS | 4.4 GB | ≈ 5.9 | ●Runs well 88–131 tok/s | 32K |
| Q3_K_Mquality drops | 4.0 GB | ≈ 5.5 | ●Runs well 94–141 tok/s | 32K |
| IQ3_XXSquality drops | 3.1 GB | ≈ 4.6 | ●Runs well 113–170 tok/s | 32K |
| Q2_Kquality drops | 3.1 GB | ≈ 4.6 | ●Runs well 115–173 tok/s | 32K |
| IQ2_XXSquality drops | 2.1 GB | ≈ 3.6 | ●Runs well 151–226 tok/s | 32K |
Click a row to choose a precision. “Maker’s file” sizes are the maker’s own files; others are parameters × bits per weight.
Compared with other hardware
Each at the highest precision it runs. Click one for details.
Smallest that runs it (Q4_K_M, 8K): GPU GeForce RTX 3080; unified memory Apple M5 · 16 GB.
Run commands
vllm serve baidu/Qianfan-VL-8B --max-model-len 8192
Memory to fine-tune
Batch 1, 2,048-token sequences, gradient checkpointing; LoRA and QLoRA train rank-16 adapters on the attention projections.
Estimates; facts from the maker’s Hugging Face repositories, as of 2026-11-06. How it’s estimated
Cited scores
From third-party leaderboards, not measured by usThis model isn’t on any leaderboard we cite yet.
Changes
All changes →No changes recorded yet.
Sources
- 百度智能云千帆：模型价格in Chinese (opens in a new tab) cloud.baidu.com · checked Oct 5, 2026
- baidu/Qianfan-VL-8B model card (opens in a new tab) huggingface.co · checked Oct 31, 2026
- baidu/Qianfan-VL-8B config.json (opens in a new tab) huggingface.co · checked Nov 3, 2026
- baidu/Qianfan-VL-8B files (opens in a new tab) huggingface.co · checked Oct 30, 2026
- 模型更新记录in Chinese (opens in a new tab) ai.baidu.com · checked Nov 5, 2026
Some values on this page are still waiting to be checked against official pages: a filled dot is verified, a hollow one is not yet (with or without a source on file). How we verify data
Baidu · China · cloud.baidu.com
Related models
Snapshots, the same family and other models from this provider