LLM API Pricing Calculator
Compare real-time API costs across Claude, GPT, Gemini, DeepSeek, and more.
Usage Calculator
| Model | Provider | Context | Input cost | Output cost | Total / call | Monthly | vs. expensive |
|---|---|---|---|---|---|---|---|
MiMo V2 Free coding model; 256K context; open weights | Xiaomi | 256K | Free | Free | Free | Free | — |
Devstral 2cheapest Cheapest agentic coding model; 256K context | Mistral | 256K | $0.000050 | $0.000110 | $0.000160 | $0.4800 | -100% |
Llama 4 Scout 10M context industry record; 109B MoE (17B active) | Meta | 10M | $0.000080 | $0.000150 | $0.000230 | $0.6900 | -100% |
DeepSeek V4 Flash Low-cost DeepSeek variant cited in 2026 token-cost comparisons as a budget option for fast inference. | DeepSeek | 0 | $0.000140 | $0.000140 | $0.000280 | $0.8400 | -99% |
Gemini 3.1 Flash Lite #3 Arena overall; #1 creative writing; ultra-fast | 1M | $0.000100 | $0.000200 | $0.000300 | $0.9000 | -99% | |
GPT-4.1 Nano Ultra-low-cost OpenAI model highlighted in 2026 pricing comparisons for lightweight agent and utility workloads. | OpenAI | 0 | $0.000100 | $0.000200 | $0.000300 | $0.9000 | -99% |
Llama 4 Maverick 400B MoE (17B active); strong multimodal; open weights | Meta | 1M | $0.000150 | $0.000300 | $0.000450 | $1.35 | -99% |
Qwen 3.6 Newer Qwen-family open model mentioned in 2026 open-source rankings for strong general-purpose performance. | Alibaba | 0 | $0.000200 | $0.000400 | $0.000600 | $1.80 | -99% |
DeepSeek V3.2 ~90% GPT-5.4 quality at 1/50th cost; best value model | DeepSeek | 128K | $0.000270 | $0.000550 | $0.000820 | $2.46 | -98% |
DeepSeek V4 A cost-focused DeepSeek release that recent leaderboards describe as a strong value option for reasoning and general use. | DeepSeek | 0 | $0.000435 | $0.000435 | $0.000870 | $2.61 | -98% |
Mistral Large 3 675B MoE (41B active); Apache 2.0; best cost-efficiency frontier | Mistral | 256K | $0.000500 | $0.000750 | $0.001250 | $3.75 | -98% |
Gemini 2.5 Flash Cheapest frontier model at scale | 1M | $0.000300 | $0.001250 | $0.001550 | $4.65 | -97% | |
DeepSeek R1 671B MoE (37B active); MIT license; distilled variants available | DeepSeek | 128K | $0.000550 | $0.001095 | $0.001645 | $4.94 | -97% |
Kimi K2 1T params; Agent Swarm (100 agents); Modified MIT | Moonshot | 128K | $0.000550 | $0.001100 | $0.001650 | $4.95 | -97% |
Kimi K2.7 Code Moonshot’s new coding-focused MoE model with a large context window and competitive pricing. | Moonshot AI | 0 | $0.000600 | $0.001200 | $0.001800 | $5.40 | -97% |
Qwen 3 235B 235B MoE (22B active); Apache 2.0; strongest OSS competitive programming | Alibaba | 128K | $0.000860 | $0.001000 | $0.001860 | $5.58 | -96% |
Claude Haiku 4.5 Fastest Claude, cheapest tier | Anthropic | 200K | $0.000800 | $0.002000 | $0.002800 | $8.40 | -95% |
GLM-5.2 Open-weights coding-focused model released in June 2026, positioned for long-horizon autonomous engineering tasks. | Z.ai | 0 | $0.001400 | $0.002200 | $0.003600 | $10.80 | -93% |
Gemini 3.5 Flash GA on May 19, 2026; faster than prior frontier models and strong on coding and agentic benchmarks. | 0 | $0.001500 | $0.004500 | $0.006000 | $18.00 | -89% | |
Gemini 2.5 Pro Thinking model; top WebDev Arena 1415; native multimodal | 1M | $0.001250 | $0.005000 | $0.006250 | $18.75 | -88% | |
GPT-4o Legacy but still available; superseded by GPT-5 family | OpenAI | 128K | $0.002500 | $0.005000 | $0.007500 | $22.50 | -86% |
GPT-5.3-Codex Coding-oriented OpenAI model referenced in June 2026 pricing coverage with strong developer-task performance. | OpenAI | 0 | $0.001750 | $0.007000 | $0.008750 | $26.25 | -83% |
GPT-5.4 Unifies Codex + GPT; 1M context; built-in computer use | OpenAI | 1M | $0.003000 | $0.007500 | $0.0105 | $31.50 | -80% |
Claude Sonnet 4.6 Best value frontier; beats Opus 4.5 in 59% head-to-head | Anthropic | 1M | $0.003000 | $0.007500 | $0.0105 | $31.50 | -80% |
Grok 3 Strong math/science; now legacy (Grok 4 series launched) | xAI | 131K | $0.003000 | $0.007500 | $0.0105 | $31.50 | -80% |
Claude Opus 4.6 #1 Arena Hard Prompts & Coding; 128K max output | Anthropic | 1M | $0.005000 | $0.0125 | $0.0175 | $52.50 | -67% |
Claude Opus 4.5 Major price cut from Opus 4; strong agentic coding | Anthropic | 200K | $0.005000 | $0.0125 | $0.0175 | $52.50 | -67% |
Grok 4 Top-5 Arena; strong reasoning & real-time X data | xAI | 256K | $0.005000 | $0.0125 | $0.0175 | $52.50 | -67% |
GPT-5.5 OpenAI's latest model, GPT-5.5, offers advanced capabilities for coding and complex tasks. | OpenAI | 0 | $0.005000 | $0.0150 | $0.0200 | $60.00 | -62% |
o3 Strongest OpenAI reasoning model | OpenAI | 200K | $0.0100 | $0.0200 | $0.0300 | $90.00 | -43% |
Claude Fable 5 Anthropic’s newly released flagship Claude model, positioned for high-end reasoning and long-context agentic work. | Anthropic | 0 | $0.0100 | $0.0250 | $0.0350 | $105 | -33% |
Claude Opus 4.8 Improves coding, reasoning, reliability, and agentic workflows while keeping standard API pricing unchanged from Opus 4.7. | Anthropic | 0 | $0.0150 | $0.0375 | $0.0525 | $158 | — |
Pricing sourced from official provider documentation. Figures are per 1 million tokens. Self-hosted models (Llama 4) cost only compute. Monthly estimate = daily calls × 30 days. Prices subject to change.