AI API Limits & Performance Matrix
Browse 34 expert guides to help you solve common ai api limits & performance matrix quickly.
Book a free consultationClaude 3 Haiku API Rate Limits & Pricing
Claude 3 Haiku deprecated March 2, 2026. Successor Haiku 4.5 limits, pricing ($1/$5/MTok), highest ITPM ceiling — verified July 2026.
Claude 3.5 Sonnet API Rate Limits & Pricing
Claude 3.5 Sonnet is legacy (dropped Feb 2026). Successor Claude Sonnet 5 limits, pricing ($2/$10/MTok intro), and 429 playbook — verified…
Claude 4 Opus API Rate Limits & Pricing
claude-opus-4 stopped June 15, 2026. Successor Opus 4.8 limits, pricing ($5/$25/MTok), 429 playbook — verified July 2026.
Claude 4 Sonnet API Rate Limits & Pricing
claude-sonnet-4 stopped June 15, 2026. Successor Sonnet 5 limits, intro pricing ($2/$10/MTok), 429 playbook — verified July 2026.
Cohere Command R API Rate Limits & Pricing
Cohere Command R (08-2024): 500 RPM production, $0.15/$0.60/MTok, full RAG stack. 429 playbook and tier guide — verified July 2026.
Devstral API Rate Limits & Pricing
Devstral API rate limits, pricing, and 429 playbook — all tiers verified July 2026. Free eval tier, pay-as-you-go baseline ~300 RPM, 256K…
Falcon 180B API Rate Limits & Pricing
Falcon 180B has no shared per-token API. License restricts third-party hosting. Only options: SageMaker JumpStart dedicated deploy…
Gemini 1.5 Flash API Rate Limits & Pricing
Gemini 1.5 Flash is shut down. See current Gemini 2.5 Flash limits ($0.30/$2.50/MTok) and how to migrate — verified July 2026.
Gemini 1.5 Pro API Rate Limits & Pricing
Gemini 1.5 Pro is shut down and retired. Verified July 2026: current limits and pricing for successor Gemini 2.5 Pro, plus migration steps.
Gemini 2.5 Flash API Rate Limits & Pricing
Gemini 2.5 Flash rate limits, pricing ($0.30/$2.50 per MTok), free tier, and 429 fixes — verified July 2026.
Gemini 2.5 Pro API Rate Limits & Pricing
Gemini 2.5 Pro rate limits, pricing ($1.25/$10 per MTok, raised July 2), 200K surcharge trap, and 429 fixes. Verified July 2026. Paid-only…
Gemma 2 API Rate Limits & Pricing
Gemma 2 has no first-party hosted API in 2026. See self-host options, Gemma 3 migration path, and limits — verified July 2026.
Gemma 3 API Rate Limits & Pricing
Gemma 3 rate limits, pricing ($0.040/$0.080 per MTok hosted), and self-host guide verified July 2026. 131K context, Apache 2.0 open weights.
GPT-3.5 Turbo API Rate Limits & Pricing
GPT-3.5 Turbo is delisted from OpenAI pricing. Legacy 429 playbook, fine-tuning deadline Jan 2027, and migration to GPT-4.1 nano — verified…
GPT-4 Turbo API Rate Limits & Pricing
GPT-4 Turbo is delisted from OpenAI's pricing table. Legacy limits, 429 playbook, and migration to GPT-5.5 or GPT-4.1 — verified July 2026.
GPT-4o API Rate Limits & Pricing
GPT-4o is now legacy. Limits (500 RPM Tier 1), Azure reference pricing ($2.50/$10/MTok), 429 playbook & migration to GPT-5.5 — verified…
GPT-5 API Rate Limits & Pricing
GPT-5 was never a distinct API endpoint. Successor GPT-5.5 limits ($5/$30/MTok), 429 playbook & migration guide — verified July 2026.
Grok 2 API Rate Limits & Pricing
Grok 2 is retired — no current first-party xAI SKU. Successor Grok 4.5 limits ($2/$6/MTok), 429 playbook, migration — verified July 2026.
Grok 4 API Rate Limits & Pricing
Grok 4 retired May 15, 2026. Successor Grok 4.5 limits (150 RPS / 50M TPM), pricing ($2/$6/MTok), 429 playbook — verified July 2026.
Imagen 4 API Rate Limits & Pricing
Imagen 4 shuts down Aug 17, 2026. Verified July 2026: per-image pricing ($0.02–$0.06), quota limits, 429 fixes, and migration guide to…
Kimi K2 API Rate Limits & Pricing
Legacy Kimi K2 EOL May 25, 2026. Successor K2.6 limits (200–5,000 RPM), pricing ($0.95/$4.00/MTok), 429 playbook — verified July 2026.
Llama 3 API Rate Limits & Pricing
Llama 3 is a legacy open-weights model. No first-party API. Get host-by-host rate limits, third-party pricing, and the migration path to…
Llama 3.1 API Rate Limits & Pricing
Llama 3.1 is a legacy model (Databricks dropped 405B Feb 2026). Host-by-host rate limits, pricing, and migration path to Llama 4 Scout…
Llama 4 Maverick API Rate Limits & Pricing
Llama 4 Maverick: host-by-host rate limits, verified pricing (DeepInfra $0.15/$0.60), 429 retry playbook, and Scout vs Maverick comparison…
Llama 4 Scout API Rate Limits & Pricing
Llama 4 Scout: host-by-host rate limits, verified pricing (DeepInfra $0.08/$0.30), 10M context explainer, 429 retry playbook. Verified July…
Mistral Large API Rate Limits & Pricing
Mistral Large 3 API rate limits, pricing, and 429 playbook — all tiers verified July 2026. $6/MTok output (verify), EU Paris routing, 128K…
Mistral Medium API Rate Limits & Pricing
Mistral Medium 3.5 API rate limits, pricing, and 429 playbook — all tiers verified July 2026. EU Paris routing, 128K context, ~300 RPM paid…
Mixtral 8x22B API Rate Limits & Pricing
Mixtral 8x22B API access paths, limits, and pricing verified July 2026. Legacy model — first-party La Plateforme status unconfirmed. Apache…
Mixtral 8x7B API Rate Limits & Pricing
Mixtral 8x7B is a legacy open-weight model. No published first-party price on La Plateforme as of July 2026. Access via Groq (free tier)…
Phi-3 API Rate Limits & Pricing
Phi-3 is superseded by Phi-4-mini on Azure AI Foundry at $0.07/$0.23 per MTok. Rate limits, pricing, and migration guide verified July 2026.
Qwen2 API Rate Limits & Pricing
Qwen2 is delisted from DashScope as of July 2026. Migration paths, third-party hosting options, and successor Qwen3-Max pricing. Verified…
Qwen3-Max API Rate Limits & Pricing
Qwen3-Max DashScope rate limits, pricing ($1.20/$6.00 per MTok), batch discount, and 429 retry playbook. Verified July 2026.
Stable Virtual Camera API Rate Limits & Pricing
Stable Virtual Camera has no hosted API as of July 2026. Access paths: Hugging Face Space demo and self-hosted open weights…
Yi-Large API Rate Limits & Pricing
Yi-Large API is closed on 01.AI; Fireworks AI is the sole active host at $3/$3 per MTok. Rate limits, pricing, and alternatives verified…