Qwen3.5-397B-A17B
Speed analysis
Latency measured across all benchmark runs. P50 (median) and P95 (95th percentile) give a realistic picture of response speed under normal and peak load.
Quality scores
Evaluation results from judge-model scoring across diverse task categories. Scores reflect coherence, accuracy and instruction-following.
Pricing history
Direct provider rates per million tokens, plus a typical-conversation cost estimate.
Pricing over time
Input & output per 1M tokens · step-line = price changes
$0.7100
input / 1M
— stable
$4.25
output / 1M
— stable
Tokens per second
Throughput in tokens per second, derived from measured P50 latency. Higher is better; fluctuations track provider-side load.
Estimated from P50 latency × 200 output tokens — the absolute number depends on this assumption; the trend is what matters.
Capabilities
Availability
Availability
How often this model answers when we call it — measured across real API requests and live tests over the last 30 days. This is separate from quality: these numbers only tell you whether the model responds, not how good the answer is.
Last 7 days
—
Last 30 days
100.0%
n=15
Median response time
1,177ms
n=15
Based on 395 measurements over the last 30 days.
Technical details
Only live API calls and live-test requests count — internal probes and benchmark runs are excluded.
Calls with a custom API key (BYOK) are excluded: those failures are key-specific, not a sign of model downtime.
Failed calls are NOT included in quality scores — quality is measured on successful responses only. Availability and quality are independent signals.
Median response time (p50) across successful calls with a recorded duration. Outliers (very slow or very fast calls) pull the median less than the average.
Total calls (30d)
15
OK responses (30d)
15
Total calls (7d)
0
OK responses (7d)
0
Tokonomix benchmark verdicts
Qwen3.5-397B-A17B jumps to 81.7/100 with creative gains, reasoning still absent
Qwen3.5-397B-A17B demonstrates a remarkable recovery with an overall quality score of 81.7, up 39.2 points from the previous window's 42.4. The model now achieves perfect scores in both coding and multilingual categories at 100 each, maintaining its strong coding performance while dramatically improving multilingual capabilities from 33. The most significant shift appears in creative tasks, which climbed from zero in the implied previous state to 45, though this remains the weakest category. However, reasoning capabilities remain completely absent with no score recorded in this window, consistent with the zero score from the previous period. Latency has increased modestly from 4725ms to 5235ms at the median, representing an approximately 11% slowdown. The test methodology remains consistent with 5 runs in each window. Users requiring strong coding and multilingual support will find this model highly capable, but those needing creative writing or reasoning tasks should be aware of the model's limitations in these areas. The dramatic quality improvement suggests either infrastructure enhancements or model configuration changes at the OVH GRA endpoint.
Quality
81.7
Latency p50
5,235 ms
Test runs
5