Cerebras
Tokenomics analysis of the Cerebras WSE-3 for inference with leading open source models. Source: SemiAnalysis Tokenomics team.
Default (96.3k) is the P50 input sequence length from our internal testing across Claude Code, Codex, Cursor, OpenCode, and Pi. Output tokens are derived from the workload mix in the Cost to serve panel below.
Where the workload mix comes from
Token-type breakdown observed across our internal workload and spend on the leading coding assistants (Claude Code, Codex, Cursor, OpenCode, Pi). Cache reads dominate by volume but cache writes and outputs disproportionately drive spend.
Where our ISL assumption comes from
Input sequence length distribution across agentic coding harnesses (Claude Code, Codex, Cursor, OpenCode, Pi). P50 lands at ~96.3k tokens.
Where our OSL assumption comes from
Output sequence length across the same harnesses. P50 lands at ~213 tokens: most turns are short replies.
Where our interactivity assumption comes from
Interactivity (output tok/s) on Cerebras: smaller models go faster, larger models go slower.
Sources: Artificial Analysis: Cerebras provider page; Cerebras: Kimi K2.6 (981 tok/s on a 1T-param model); Kimi API model list (256k context capacity).
Fit to interactivity ≈ 3322 / active_params_B0.294 using gpt-oss 120B (2059) and GLM 4.7 (1201) as anchors. Kimi K2.6 uses the Cerebras-reported 981 tok/s measurement directly. Models with compressed, sparse, or linear attention (DeepSeek V3/V4, Kimi K3, MiniMax M3, GLM 5.2) are bumped slightly above the curve to reflect their smaller per-step KV bandwidth.
| Model | Active params | Interactivity (tok / sec) |
|---|---|---|
| DeepSeek V4 Pro | 49 B | 1,300 |
| DeepSeek V4 Flash | 13 B | 1,900 |
| Kimi K3 | 104 B | 1,050 |
| MiniMax M3 | 23 B | 1,400 |
| GLM 5.2 | 40 B | 1,250 |
| Kimi K2.6 | 32 B | 981 |
| gpt-oss 120B | 5.1 B | 2,059 |
| GLM 4.7 | 32 B | 1,201 |
| DeepSeek V3 | 37 B | 1,350 |
Cerebras WSE-3 vs other chips
Comparing chip and system specs.