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
How this model compares to the rest of the field on each prompt category, from a pairwise fit over the same prompts. The raw judge score sits underneath each number.
Win rate per category: how often this model beats a field-average model on a prompt from that category. 50% is average, not a failing grade. It is not a percentage of correct answers.
Pricing
What you pay per million tokens when you use this model on Tokonomix, plus an estimate for a typical conversation.
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
100.0%
n=386
Last 30 days
100.0%
n=1,911
Median response time
4,415ms
n=1,911
Based on 2,291 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)
1,911
OK responses (30d)
1,911
Total calls (7d)
386
OK responses (7d)
386
Tokonomix benchmark verdicts
Quality reaches 92.8 as reasoning and factual hit perfect scores
Meta-Llama-3.3-70B-Instruct delivered a strong performance improvement, climbing 5.8 points to reach an overall quality score of 92.8. The model achieved perfect scores in both reasoning and factual categories, each hitting 100 points, demonstrating exceptional capabilities in logical processing and information accuracy. This represents a significant advancement in factual understanding, which previously scored only 85. Multilingual performance showed solid improvement, rising from 77 to 88 points, indicating better language handling across diverse linguistic contexts. However, coding performance declined from a perfect 100 to 83 points, representing the most notable regression in this benchmark window. Latency increased slightly from 12076ms to 13286ms at the median, representing approximately a 10% slowdown in response times. While this increase is measurable, the model still maintains reasonable performance for a 70B parameter model. The overall trajectory remains positive with the quality score continuing its upward trend across consecutive benchmark windows. Users can expect highly reliable reasoning and factual outputs, strong multilingual support, and competent coding assistance, though the latter has stepped back from its previous peak performance. The model demonstrates particular strength in knowledge-intensive and analytical tasks.
Quality
92.8
Latency p50
13,286 ms
Test runs
5