Unverified50% confidenceBenchmarkExact time
In an ultra-long context test, Ornith 1.5 ran at around 75 tokens/second early on and dropped to approximately 45 tokens/second after exceeding 1.6 million tokens, while Qwen 3.8 dropped to approximately 5 tokens/second under the same load.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/11/2026
First Seen
Valid until: 12/10/2026
Sources
Related Entities
Related Claims
UnverifiedUnder a TP4 tensor-parallel configuration, QWEN 3.8 can peak at 100–115 tokens per second.66% similarUnverifiedQwen 3.8 has a base speed of 22 tokens/s on the M5 Max, which can be boosted to 34 or even 56 tokens/s with MTP enabled.65% similarUnverifiedIn speculative sampling, when the acceptance rate of candidate tokens is high, the effective number of output tokens per forward pass can reach 2–4, thereby improving the overall generation speed by 1.5x–3x.61% similarUnverifiedIn token generation tests for the Gemma 4 12B model, Spark reached 26.4 tokens/s, Halo 24.6 tokens/s, and M4 Pro was the fastest at 33.8 tokens/s.59% similarUnverifiedOrnith 35B MoE provided the correct answer to the Three Gods Problem in approximately 30 seconds, while both Ornith 9B and the original Qwen 3.5-9B failed.59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/907313API
curl https://kongchang.com/api/v1/knowledge/claims/907313MCP
get_claim(id=907313)