The Token Ledger Digest – 2026-06-01
Most cost‑impacting change
- inclusionAI: Ring-2.6-1T – Prompt price rose from $0.075/1M to $0.30/1M (4×); completion price rose from $0.625/1M to $2.50/1M (4×). Who should care: Teams running high‑volume inference on this model will see a four‑fold increase in per‑token cost; consider re‑evaluating usage or switching to alternatives.
Added models
- MiniMax: MiniMax M3 – Prompt $0.30/1M, completion $1.20/1M, 1M‑token context. Who should care: Applications needing very long context (up to 1M tokens) with moderate pricing.
- Upstage: Solar Pro 3 – Prompt $0.15/1M, completion $0.60/1M, 128k‑token context. Who should care: Users seeking a cheaper mid‑range model with decent context length.
Removed models (no longer available)
- DeepSeek V4 Flash (free) – $0/$0
- Xiaomi MiMo‑V2‑Omni – $0.40/1M prompt, $2.00/1M completion
- Xiaomi MiMo‑V2‑Pro – $1.00/1M prompt, $3.00/1M completion
- Mistral Devstral Medium – $0.40/1M prompt, $2.00/1M completion
- Mistral Devstral Small 1.1 – $0.10/1M prompt, $0.30/1M completion
- Mistral Large 2411 – $2.00/1M prompt, $6.00/1M completion
- Mistral Pixtral Large 2411 – $2.00/1M prompt, $6.00/1M completion Who should care: Migrate any workloads off these IDs to avoid broken calls.
Other price adjustments
- Tencent: Hy3 preview – Prompt $0.063→$0.066/1M (+5%); completion $0.21→$0.26/1M (+24%).
- Z.ai: GLM 5 – Prompt unchanged $0.60/1M; completion $1.92→$2.08/1M (+8%). Who should care: Minor cost uplift; monitor if usage scales significantly.
Total models tracked: 345.
Originally published at The Token Ledger. Subscribe for the daily digest.
Top comments (0)