Running Qwen3.5 on B300
Quick answer
Qwen3.5 sustains 15,815 tokens/s per GPU on B300 at 50 tokens/s per user, which works out to $0.040 per million tokens at hyperscaler pricing, served by SGLang. Fastest measured TTFT: 0.1 ms; fastest TPOT: 0.0 ms (each the best across all configs, not one run).
Benchmarked configs
178
Serving engines
SGLang
Precisions
bf16, fp4, fp8
Run dates
2026-05-18 → 2026-08-10
Throughput at every interactivity target
Serving is a trade-off: push more concurrent users through a GPU and each user's tokens arrive slower. The ladder below reads the measured frontier at each per-user speed target on a single-turn chat workload (8k input / 1k output), using the best engine and precision at that point.
| Per-user target | Tokens/s per GPU | $ / 1M tokens | Engine | Precision |
|---|---|---|---|---|
| 30 tok/s | 9,053 | $0.069 | SGLang | fp8 |
| 50 tok/s | 15,815 | $0.040 | SGLang | fp4 |
| 75 tok/s | 12,650 | $0.050 | SGLang | fp4 |
| 100 tok/s | 10,337 | $0.061 | SGLang | fp4 |
| 150 tok/s | 7,165 | $0.088 | SGLang | fp4 |
| 200 tok/s | 5,286 | $0.12 | SGLang | fp4 |
What serving actually costs
Converting the 50 tokens/s per user operating point to $ per million total tokens across rental pricing tiers from the SemiAnalysis AI Cloud TCO model.
| Pricing tier | $ / GPU / hr | $ / 1M tokens |
|---|---|---|
| Hyperscaler | $2.26 | $0.040 |
| Neocloud | $2.52 | $0.044 |
| Retail | $3.00 | $0.053 |
Frequently asked questions
- How fast is Qwen3.5 on B300?
- At an interactivity target of 50 tokens/s per user on a single-turn chat workload (8k input / 1k output), B300 sustains 15,815 tokens/s per GPU serving Qwen3.5 with SGLang in FP4. Peak measured throughput across all configs is 97,865 tokens/s per GPU.
- How much does it cost to serve Qwen3.5 on B300?
- $0.040 per million total tokens at hyperscaler $/GPU/hr pricing, at 50 tokens/s per user. Neocloud and retail rental tiers are tabulated above; slower interactivity targets lower the cost further.
- Which serving engines run Qwen3.5 on B300?
- The runs behind this page used SGLang in BF16, FP4, FP8. Engines are rebuilt and re-benchmarked continuously, so the best config can change between visits.
- How are these Qwen3.5 numbers measured?
- Every number is measured on real B300 hardware by the InferenceX fleet, sweeping concurrency on a single-turn chat workload (8k input / 1k output) to trace the throughput-versus-interactivity frontier; the newest run landed on 2026-08-10. The same derivation powers the InferenceX overview leaderboard.