AI model weekly: Week ending October 11, 2026 — Ling 3.0 Flash Sante lands

Developers should prioritize reliability metrics as OpenAI and Groq report significant outages, while monitoring price volatility in DeepSeek and Kimi models. New entries from Anthropic and Mistral expand context window options, contrasting with a wave of Qwen and MiMo deprecations scheduled for October 2026.

Outages & incidents

• Fireworks AI — Service Degradation for GLM 5.3 (impact: minor) · status
• Fireworks AI — Service Degradation for Kimi K3 Fast (impact: minor) · status
• Fireworks AI — Service Degradation for Qwen 3.8 Max (impact: minor) · status
• Groq — Model Performance or Availability Issue - minimaxai/minimax-m2.7 (impact: major) · status
• Fireworks AI — Service Degradation for Qwen 3.8 Max (impact: minor) · status
• HeyGen — incident fixed and monitoring · status
• Vercel / v0 — Delays in Drains, Observability, and Workflow in Washington (iad1) (impact: major) · status
• Runway — We've identified a problem and are working on a fix (impact: major) · status
• Perplexity — Elevated API errors and timeouts (impact: minor) · status
• Cursor — Investigating service degradation — Cloud Agents and Grok Bot (impact: minor) · status
• OpenAI / ChatGPT — Workspace admins may be unable to manage plugins (impact: major) · status
• Cursor — Investigating service degradation — Cloud Agents (impact: minor) · status

Reliability watch

• OpenAI / ChatGPT: 293 failed and 2600 degraded checks out of 5033
• Groq: 183 failed and 130 degraded checks out of 5033
• ElevenLabs: 64 failed and 93 degraded checks out of 5033
• Lovable: 52 failed and 49 degraded checks out of 5033
• SambaNova: 5 failed and 13 degraded checks out of 5033

Price changes

• DeepSeek V4 Flash 0731 (DeepSeek): $0.0161 / $0.64 → $0.0122 / $0.90 (+41%)
• GLM 5.3 (Z.ai (Zhipu)): $0.07 / $4.80 → $0.079 / $4.80 (0%)
• DeepSeek V4 Flash 0731 (DeepSeek): $0.0175 / $0.64 → $0.0161 / $0.64 (0%)
• Kimi K3 (Moonshot AI): $0.80 / $13.50 → $0.73 / $13.50 (0%)
• Kimi K3 (Moonshot AI): $0.80 / $10.00 → $0.80 / $13.50 (+35%)
• Hy3 (Tencent): $0.13 / $0.53 → $0.0825 / $0.33 (-37%)
• Hy4 preview (Tencent): $0.83 / $2.50 → $0.75 / $2.25 (-10%)
• GLM 5.3 (Z.ai (Zhipu)): $0.039 / $4.80 → $0.07 / $4.80 (0%)
• GLM 5.3 (Z.ai (Zhipu)): $0.07 / $4.80 → $0.039 / $4.80 (0%)
• DeepSeek V4 Flash 0731 (DeepSeek): $0.0138 / $0.64 → $0.0175 / $0.64 (0%)
• GLM 5.3 (Z.ai (Zhipu)): $0.039 / $4.80 → $0.07 / $4.80 (0%)
• Nemotron 3 Ultra (NVIDIA): $0.50 / $2.20 → $0.60 / $2.40 (+9%)
• DeepSeek V4 Flash 0731 (DeepSeek): $0.018 / $0.64 → $0.0138 / $0.64 (0%)
• GLM 5.3 (Z.ai (Zhipu)): $0.07 / $4.80 → $0.039 / $4.80 (0%)
• DeepSeek V4 Flash 0731 (DeepSeek): $0.0161 / $0.64 → $0.018 / $0.64 (0%)

New models

• Ling 3.0 Flash Sante — inclusionAI · $0.042 / $0.12 · 262K ctx
• Step 5 Preview — StepFun · $1.00 / $2.70 · 1000K ctx
• Claude Haiku 5.5 — Anthropic · $0.10 / $0.50 · 1000K ctx
• Nano Banana 2.1 — Google · $1.50 / $7.50 · 66K ctx
• Mistral Large 4 — Mistral AI · $0.68 / $2.09 · 1049K ctx

Deprecations & removals

• Space Bunny Alpha retires on 2026-10-05
• Qwen3 Next 80B A3B Instruct retires on 2026-10-08
• Qwen3.5-9B retires on 2026-10-21
• MiMo-V2.5-Pro retires on 2026-10-21
• MiMo-V2.5 retires on 2026-10-21
• Space Bunny Alpha was removed from the catalog
• KAT-Coder-Pro V2.5 was removed from the catalog
• ERNIE 4.5 VL 424B A47B was removed from the catalog
• Qwen3.6 Max Preview was removed from the catalog
• Qwen3 Max Thinking was removed from the catalog
• Qwen3 VL 32B Instruct was removed from the catalog
• Qwen3 VL 8B Thinking was removed from the catalog
• Qwen3 VL 235B A22B Thinking was removed from the catalog
• Qwen3 Max was removed from the catalog
• Qwen3 Coder Plus was removed from the catalog
• Qwen Plus 0728 was removed from the catalog
• Qwen3 30B A3B Thinking 2507 was removed from the catalog

This brief is generated automatically from ModelBank's live monitoring data. Want these changes the moment they happen? Get free alerts.