Quick reference: current text-model list prices
| Model | Status | Input | Cached input | Output | Context / pricing breakpoint | Use it first for | Primary source |
|---|---|---|---|---|---|---|---|
| OpenAI GPT-6 Astra | Available | $10.00 | $1.00 | $50.00 | 1.05M; >272K prompt is 2× input, 1.5× output | Hardest end-to-end reasoning, coding, computer use | OpenAI model card |
| OpenAI GPT-6 Sol | Available | $2.00 | $0.20 | $10.00 | 1.05M; >272K prompt is 2× input and cache | Complex coding and agentic workflows | OpenAI model card |
| OpenAI GPT-6 Luna | Available | $0.10 | $0.01 | $0.50 | 1.05M; >272K prompt is 2× input and cache | Focused, high-volume extraction and summarization | OpenAI model card |
| OpenAI GPT-6.1 Terra | Listed on OpenAI pricing page | $1.00 | $0.05 | $5.00 | Not verified; prompts over 272K are 2× input and cache, 1.5× output | Lower-cost GPT-6.1 tier; use case not verified | OpenAI pricing |
| Anthropic Claude Fable 5.1 | Available | $10.00 | $0.25 | $50.00 | 1M standard pricing | Most demanding reasoning and long-horizon agents | Anthropic pricing |
| Anthropic Claude Opus 5.5 | Available | $4.00 | $0.20 | $20.00 | 1M standard pricing | High-capability reasoning/coding | Anthropic pricing |
| Anthropic Claude Sonnet 5.5 | Available | $2.00 | $0.10 | $10.00 | 1M standard pricing | Balanced agentic workloads | Anthropic pricing |
| Anthropic Claude Haiku 5.5 | Available | $0.10 (≤100K prompts); $0.50 over | $0.01 (≤100K); $0.05 over | $0.50 (≤100K); $2.50 over | Two prompt-size tiers at 100K tokens | Fast routing/classification | Anthropic pricing |
| Anthropic Claude Haiku 4.5 | Available | $1.00 | $0.10 | $5.00 | Read exact model context terms | Fast routing/classification | Anthropic pricing |
| Google Gemini 4 Argon | Available; promo to Nov 30, 2026 | $2.00 (then $4.00) | $0.10 (then $0.20) | $10.00 (then $20.00) | Not verified; caching is implicit only | Newest Gemini text model; use case not verified | Gemini pricing |
| Google Gemini 3.6 Flash | Available; promo to Dec 31, 2026 | $0.75 (then $1.50) | $0.075 (then $0.15) | $3.75 (then $7.50) | Context-cache storage fees apply | Flash-tier workloads; use case not verified | Gemini pricing |
| Google Gemini 3.8 Flash | Available; promo to Dec 31, 2026 | $0.75 (then $1.50) | $0.075 (then $0.15) | $3.75 (then $7.50) | Tool/search charges can apply | Long-horizon coding, agents, enterprise workflows | Gemini pricing |
| Google Gemini 3.5 Flash-Lite | Available | $0.30 | $0.03 | $2.50 | Tools billed separately where used | High-volume, simple processing | Gemini pricing |
| Google Gemini 3.1 Flash-Lite | Available | $0.25 | $0.025 | $1.50 | Tools billed separately where used | Cheapest multimodal batch-style work | Gemini pricing |
| Google Gemini 3.1 Pro | Preview | $2.00 ≤200K | $0.20 ≤200K | $12.00 ≤200K | Input/output step up above 200K ($4 / $18) | Multimodal/agent evaluation | Gemini pricing |
| Mistral Medium 3.5 | Available; open weights | $1.50 | See live card | $7.50 | 256K | Agentic and coding use, self-hostable | Mistral model card |
| Mistral Small 4 | Available; open weights | $0.15 | See live card | $0.60 | 256K | Low-cost hybrid reasoning/coding | Mistral model card |
The table compares standard text rates, not Claude Mythos 5.1 (invitation-only, $10 / $50), older Claude (Opus 5 and 4.8 at $5 / $25, Sonnet 5 at $2 / $10, now permanent), enterprise agreements, regional premiums, priority/flex, audio/image/video, storage, or third-party cloud resale.
Workload cost calculator
Cost vocabulary and a worked estimate
What can be billable?
- Input tokens: system instructions, user content, retrieved text, tool schemas, and conversation history.
- Output and reasoning tokens: generated text and, depending on the provider/model, hidden or intermediate reasoning consumed by an agent loop.
- Cached input: a reused prefix; cache reads are cheaper, while writes/storage can cost more than a normal input token.
- Tools and modalities: web search, grounding, code execution, containers, images, audio, video, retrieval, and storage can have separate meters.
- Quota: RPM, TPM, concurrent requests, and batch queue size can make an otherwise cheap row unusable.
- Resilience: retries, fallbacks, duplicate tool calls, and human review are deployment costs—not free edge cases.
GPT-6 Sol planning example
10,000 requests × 1,800 input + 450 output, with half of input read from cache:
cached input = 9.0M × $0.20 = $1.80
output = 4.5M × $10.00 = $45.00
token estimate = $64.80 / month
This excludes tools and cache writes. For >272K GPT-6 prompts, recalculate at the published 2× input and cache multipliers (Astra also 1.5× output)—not the base row.
Choose the operating mode before the model
| Constraint | Start with | Why | Do not assume |
|---|---|---|---|
| Hardest reasoning and agent work, cost secondary | GPT-6 Astra or Claude Fable 5.1 | Both list at $10 / $50 with 1M-class context | That the flagship tier is needed for every step of an agent loop |
| Default coding and agent workhorse | GPT-6 Sol, Claude Sonnet 5.5, or Claude Opus 5.5 | $2 / $10 for Sol and Sonnet 5.5; Opus 5.5 at $4 / $20 with a 0.05× cache-read rate | That equal list price means equal tokens per task; Claude 4.7+ tokenizers emit about 30% more tokens |
| High-volume simple routing | GPT-6 Luna, Gemini 3.1 Flash-Lite, or Gemini 3.5 Flash-Lite | Luna lists at $0.10 / $0.50; the Gemini Lite rows are $0.25–0.30 in | That quality, tool reliability, or quota are interchangeable |
| Cheap frontier-adjacent Flash tier | Gemini 3.8 Flash | $0.75 / $3.75 through Dec 31, 2026 | That the promo rate survives into 2027; budget $1.50 / $7.50 |
| Cheaper Claude for classification | Claude Haiku 5.5 | $0.10 / $0.50 for prompts up to 100K (batch $0.05 / $0.25); Haiku 4.5 at $1 / $5 is still listed | That older tokenizer measurements still predict cost |
| Open-weight control | Mistral Medium 3.5, Mistral Small 4, or another evaluated open family | Weights may lower lock-in and enable self-hosting | That downloads eliminate GPU, ops, licence, or security cost |
Advanced cost traps and anti-patterns
Preview, promo, and fallback design
A price can be temporary or ahead of access: GPT-5.6 Sol, Terra and Luna were limited-preview rows until GPT-6 replaced them at half the price, Gemini 3.8 Flash doubles on Jan 1, 2027, and Gemini 3.1 Pro is still a preview. Pin a dated model identifier where the provider supports snapshots, keep a tested fallback in the same product path, and make your rate/quota behavior observable. Consumer plans are not API credits.
Context, caching, and agent loops
Long context is capacity, not a mandate to send the corpus. Cache-read rates are not uniform: Claude reads are 0.1× input, but 0.05× on Opus 5.5 and Sonnet 5.5, and 0.025× on Fable 5.1, while GPT-6 Astra, Sol and Luna read at 0.1×. Cache a stable prefix only when the provider’s cache lifetime, write charge, and hit behavior work for your traffic. Cap agent turns, tool calls, retrieved documents, and output tokens; intermediate work can dominate the final answer.
Apples-to-apples evaluation
Price per token is not price per completed task. Compare models on identical prompts, tool schemas, retry rules, safety policy, output limits, region, and service tier. Re-tokenize when a vendor changes tokenizer: Anthropic documents that newer Claude models can produce about 30% more tokens for the same text.
Common mistakes
- Using a subscription price or an outdated screenshot as API pricing evidence.
- Comparing input-only price while ignoring output, cache writes, tools, and storage.
- Budgeting a promo price (Gemini 3.8 Flash) past its end date, or a preview model with no fallback.
- Assuming published rate limits are sustained throughput guarantees.
- Using a giant context to avoid retrieval, ranking, or document selection.
- Calling self-hosting “free” because the weights are downloadable.
Continue the decision
Frontier AI labs · Open-weight model deployment · AI model picker · AI coding agents compared.
Sources: OpenAI models · OpenAI pricing · GPT-6 Sol and Luna launch (Sep 22, 2026) · Anthropic pricing · Gemini pricing · Mistral Medium 3.5 · Mistral Small 4.