Qwen: Qwen3.8 27B
qwen/qwen3-8-27bQwen3.8 27B is an open-weight dense vision-language model from Qwen.
Modalities
In / out per 1M$0.24 / $2.20
Context1M
ReleasedAug 14, 2026
Providers
20 providers serve this model upstream.
| Provider | Input /M | Output /M | Cache read /M | Latency P50 | Throughput P50 | Uptime |
|---|---|---|---|---|---|---|
 ModelRun [by Modular]fp4 | $0.20 | $2.50 | $0.09 | 0.39s | 192 tps | 100% |
| $0.04 | $2.50 | $0.02 | 0.75s | 97 tps | 100% | |
 NEAR AIfp8 | $0.04 | $1.35 | $0.02 | 1.33s | 64 tps | 100% |
| $0.24 | $2.20 | $0.05 | 1.08s | 46 tps | 100% | |
| $0.45 | $3.20 | Variable | 1.12s | 71 tps | 100% | |
| $0.04 | $2.20 | $0.02 | 0.38s | 25 tps | 100% | |
| $0.23 | $1.98 | $0.05 | 0.47s | 52 tps | 100% | |
| $0.99 | $1.49 | $0.99 | 0.34s | 561 tps | 100% | |
 DeepInfraus · bf16 | $0.15 | $1.88 | $0.04 | 1.41s | 37 tps | 99.97% |
| $0.09 | $2.35 | $0.09 | 0.51s | 65 tps | 99.89% | |
| $0.42 | $3.00 | $0.09 | 1.30s | 40 tps | 99.85% | |
| $0.04 | $1.25 | $0.00 | 2.60s | 30 tps | 99.83% | |
| $0.04 | $2.30 | $0.04 | 0.56s | 106 tps | 99.82% | |
| $0.45 | $3.20 | $0.05 | 0.38s | 34 tps | 99.28% | |
| $0.02 | $0.50 | $0.01 | 3.73s | 5 tps | 99.18% | |
| $0.20 | $2.50 | Variable | 0.41s | 56 tps | 99.11% | |
| $0.20 | $2.50 | $0.05 | 0.81s | 80 tps | 98.86% | |
| $0.42 | $2.55 | $0.09 | 0.60s | 47 tps | 98.81% | |
 Darkbloomfp4 | $0.04 | $2.20 | $0.02 | 0.98s | 55 tps | 93.88% |
 CoreWeavefp8 | $0.40 | $3.00 | $0.15 | 0.29s | 28 tps | 30.81% |
Market-wide measurements, refreshed every 15 minutes. They describe who serves Qwen3.8 27B in general, not how ProjectCOZY routes your requests.
Pricing
List rates per 1M tokens — what Qwen3.8 27B costs pay-as-you-go elsewhere, not what you pay here.
Input
$0.24
per 1M tokens
Output
$2.20
per 1M tokens
Cache read
$0.05
per 1M cached tokens
Included in every pass
Qwen3.8 27B is included in every ProjectCOZY pass at no extra cost — one flat rate, no per-token metering. The rates above are there for comparison.
Performance
Time to first token and output speed across every provider, over the last 3 days.
0.0s50.0s100.0s150.0s200.0s
Oct 16, 12 AMOct 16, 8 AMOct 16, 4 PMOct 17, 12 AMOct 17, 8 AMOct 17, 4 PMOct 18, 12 AMOct 18, 8 AMOct 18, 4 PMOct 19, 12 AMOct 19, 8 AM
Tenstorrent Reka AI AkashML DekaLLM NEAR AI Mancer NovitaAI Phala Darkbloom Wafer Parasail Cerebras Chutes CoreWeave Venice DeepInfra (us) ModelRun [by Modular] Ionstream Alibaba Cloud Int. Cloudflare
0 tok/s1000 tok/s2000 tok/s3000 tok/s4000 tok/s
Oct 16, 12 AMOct 16, 8 AMOct 16, 4 PMOct 17, 12 AMOct 17, 8 AMOct 17, 4 PMOct 18, 12 AMOct 18, 8 AMOct 18, 4 PMOct 19, 12 AMOct 19, 8 AM
Tenstorrent Reka AI AkashML DekaLLM NEAR AI Mancer NovitaAI Phala Darkbloom Wafer Parasail Cerebras Chutes CoreWeave Venice DeepInfra (us) ModelRun [by Modular] Ionstream Alibaba Cloud Int. Cloudflare
Uptime
100% of requests were served successfully over the last 3 days, across all providers.
Last 3 daysAvailability 99.72%
Oct 16, 12:00 PMNow
Availability over the last 24 hoursAvailability 99.81%
100% 98% 97% 95%
Oct 18, 12:20 PMNow
Apps
Tools that most often run Qwen3.8 27B, by tokens this week.
Activity
1785.3B tokens processed market-wide in the last month.
050.0B100.0B150.0B200.0B
Sep 19Sep 24Sep 29Oct 4Oct 9Oct 14Oct 19
Tokens
FAQ
Which model id do I use for Qwen3.8 27B?
qwen/qwen3.8-27b. Send it as the model field from any OpenAI-compatible client, against http://192.168.1.55/v1.
How much context does Qwen3.8 27B hold?
1,000,000 tokens per request, of which up to 235,929 can be the response.
What can it accept and return?
Input: text, image, video. Output: text.
What does Qwen3.8 27B cost on a pass?
Nothing beyond the pass itself. Pay-as-you-go elsewhere it lists at $0.24 per 1M input tokens and $2.20 per 1M output tokens.
Who built Qwen3.8 27B?
Qwen, in the Qwen3.8 family.
Explore
Other models your pass can call.
