⠿ swyrmcomputeBack to website
Models

Model and token pricing.

Model / API IDContextInput / 1MCached / 1MOutput / 1M
Qwen3.8-Flash-Nextswyrm/qwen3.8-flash-nextUp to 262K tokens, certified per node$0.12$0.03$0.45

Measured speed

About 50 output tokens per second and 450–490 prompt tokens per second. A 32K-token prompt takes about a minute the first time and about 0.2 seconds on a follow-up that reuses the conversation cache. Larger GPUs are faster. Every node is benchmarked and certified for the context it can really serve (32K, 128K or 262K).

Best uses

  • Long documents processed in the background (summaries, extraction, review).
  • Long conversations and coding sessions, where the cache makes follow-ups fast and cheap.