•Qwen3.8 27B•

Open weights model

•

Released August 2026

Qwen3.8 27B (Xhigh) API Provider Benchmarking & Analysis

Model Comparison

This analysis is intended to support you in choosing the best API provider of Qwen3.8 27B (Xhigh) for your use-case.

Lowest Time per Task

#1
Nebius (FP4)Nebius (FP4)
5.6 min
#2
CrusoeCrusoe
6.7 min
#3
Multiverse ComputingMultiverse Computing
7.3 min
#4
Modular (NVFP4)Modular (NVFP4)
8.2 min
#5
WaferWafer
10.8 min

Time per Intelligence Index task (minutes)

Total 10 providers

Lowest Cost per Task

#1
Modular (NVFP4)Modular (NVFP4)
$0.00
#2
WaferWafer
$1.05
#3
Multiverse ComputingMultiverse Computing
$1.47
#4
DeepInfraDeepInfra
$1.54
#5
Parasail (FP8)Parasail (FP8)
$1.78

Cost per Intelligence Index task

Total 10 providers

Qwen3.8 27B (xhigh) is available through 10 API providers, each offering different performance characteristics and pricing. Below is a comparison of the key metrics across providers.

  • For time per task, Nebius (FP4) (5.6 min), Crusoe (6.7 min), and Multiverse Computing (7.3 min) finish an Intelligence Index task fastest.
  • For cost per task, Modular (NVFP4) ($0.00), Wafer ($1.05), and Multiverse Computing ($1.47) are the cheapest per Intelligence Index task.
Inference time (minutes) per Intelligence Index task · Lower is better
Weighted average cost (USD) per Intelligence Index task · Lower is better

Pricing

Cost per Intelligence Index Task: Qwen3.8 27B (Xhigh)

Weighted average cost (USD) per Intelligence Index task · Segmented by token type · Lower is better

Pricing: Cache Hit, Input, and Output: Qwen3.8 27B (Xhigh)

Price (USD per M Tokens) · Lower is better

Output Speed vs. Cost per Task: Qwen3.8 27B (Xhigh)

Output speed: output tokens divided by total request time, including time to first token · Weighted average cost (USD) per Intelligence Index task
Most attractive quadrant
Pareto line

Speed

Average of the single-request and multi-turn agentic workloads measured

Time per Intelligence Index Task: Qwen3.8 27B (Xhigh)

Output tokens divided by end-to-end output speed (minutes per Intelligence Index task) · Lower is better

Latency

Average of the single-request and multi-turn agentic workloads measured

Time to First Answer Token: Qwen3.8 27B (Xhigh)

Seconds until the first answer token after any reasoning · Average of the single-request and multi-turn agentic workloads measured · Lower is better

Reliability

Share of requests the endpoint failed, across all the single-request and multi-turn agentic workloads measured

Endpoint Error Rate: Qwen3.8 27B (Xhigh)

Share of requests the endpoint failed with an error, a rate limit or a timeout · Across all the single-request and multi-turn agentic workloads measured · Lower is better

Caching

Cache Hit Rate: Qwen3.8 27B (Xhigh)

Median across multi-turn agentic session runs in the last four weeks, updated Oct 26, 2026 · Higher is better

Cost per Task vs. Cache Hit Rate: Qwen3.8 27B (Xhigh)

Weighted average cost (USD) per Intelligence Index task · Cache hit rate: median of the last four weeks, updated Oct 26, 2026
Most attractive quadrant
Pareto line

Key Comparison Metrics & API Features

Alibaba Cloud logoAlibaba Cloud
$3.53
40.8
126.25
0.9%
1M
Open
CoreWeave logoCoreWeave
$3.30
27.0
130.45
2.3%
262k
Open
Crusoe logoCrusoe
$2.61
6.8
23.57
0.0%
262k
Open
DeepInfra logoDeepInfra
$1.54
36.5
131.95
0.0%
262k
Open
Modular logoModular
$0.00
8.6
26.73
0.7%
262k
Open
Multiverse Computing logoMultiverse Computing
$1.47
8.0
22.63
0.7%
262k
Open
Nebius logoNebius
$1.98
5.6
17.01
0.9%
262k
Open
Parasail logoParasail
$1.78
24.5
76.33
2.4%
262k
Open
SiliconFlow logoSiliconFlow
$2.73
89.3
241.53
21.7%
262k
Open
Wafer logoWafer
$1.05
12.9
54.67
5.9%
262k
Open

Frequently Asked Questions

Common questions about Qwen3.8 27B (Xhigh) providers