Unverified50% confidenceBenchmarkExact time
The framework's prefill speed cap is approximately 900 to 1,000 tokens, and generally hovers around 900 on Windows after resource contention.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/5/2026
First Seen
Valid until: 1/3/2027
Sources
Related Entities
Related Claims
UnverifiedGrok 4.5 outputs at 80 tokens per second and features a context window of up to 500,000 tokens (approximately 400,000 English words).72% similarUnverifiedIn prefill speed tests, Halo achieved approximately 650 tokens/s while Spark achieved approximately 2000 tokens/s, with Spark being nearly three times faster.68% similarUnverifiedHarness has approximately 900+ dependencies totaling about 1 GB, and the initial installation typically takes 2 to 5 minutes.66% similarUnverifiedUnder this configuration, the 35B model achieves a measured single-user peak speed of approximately 80 tokens/second, with concurrent throughput exceeding 500 tokens/second.65% similarUnverifiedUltrafast can generate content at a speed of up to 750 tokens per second.64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/978819API
curl https://kongchang.com/api/v1/knowledge/claims/978819MCP
get_claim(id=978819)