Prices and official documentation checked on 15 October 2026. All dollar amounts below are USD. Local prices, taxes, promotions and account eligibility can differ; check your checkout screen before subscribing.
A monthly subscription buys access and an included usage allowance. It does not necessarily buy unlimited work, and it does not automatically fund a developer API account. Start with the subscription tables, then use the model tables only if you need metered usage.
Three terms that make the prices easier to understand
- Plan: the recurring subscription, such as Plus or Pro.
- Token: a unit of model input or output. A message can contain many tokens, including files and conversation history.
- Credit: a provider-specific billing unit. One GitHub credit is not interchangeable with one ChatGPT or Google credit.
“Input” means material processed by the model. “Output” means generated material, including billable reasoning where the provider specifies it. An agent task can make several model calls, so one task is not necessarily one charge.
1. ChatGPT: subscription access and metered credits
| Plan | Monthly reference price |
|---|---|
| Free | $0 |
| Go | $8 |
| Plus | $20 |
| Pro | $100 / $200 / $500 tiers |
| Business | $25/user monthly; $20/user/month billed annually, 2+ users |
| Enterprise / Edu | Contact sales |
Source: official OpenAI documentation. Below are Work/Codex Standard credits, not ordinary Chat message prices. Included limits are separate. GPT-6.1 Sol requires Work/Codex access.
| Model or feature | Input / cached input / output credits per 1M tokens |
|---|---|
| GPT-6 Astra | 250 / 25 / 1,250 |
| GPT-6.1 Sol | 50 / 2.5 / 250 |
| GPT-6 Sol | 50 / 5 / 250 |
| GPT-6 Luna | 2.5 / 0.25 / 12.5 |
| GPT-5.6 Sol / Daybreak Blue | 100 / 10 / 500 |
| GPT-5.6 Terra | 50 / 5 / 300 |
| GPT-5.6 Luna | 5 / 0.5 / 30 |
| Daybreak Red | 312.5 / 31.25 / 1,875 |
| GPT-Rosalind-Research | 125 / 12.5 / 625 |
| GPT-Image-2, image tokens | 200 / 50 / 750 |
| GPT-Image-2, text tokens | 125 / 31.25 / 250 |
Specialist eligibility and speed affect access and cost. Credit purchase prices depend on your plan.
For developers, OpenAI API pricing is a separate dollar rate card. Current flagship Standard rates for prompts up to 272K input tokens are:
| API model | Input / output per 1M tokens |
|---|---|
| GPT-6 Astra | $10 / $50 |
| GPT-6.1 Sol | $2 / $10 |
| GPT-6 Luna | $0.10 / $0.50 |
Longer prompts, caching, tools and faster processing have separate rates. The linked API card also covers older and specialist models. Do not use API prices to estimate how many tasks a subscription includes.
2. Gemini: Google AI plans and product-specific usage
| Plan | Published US monthly reference |
|---|---|
| Free Gemini access | $0 |
| Google AI Plus | $7.99 US launch reference; verify current local offer |
| Google AI Pro | $19.99 |
| Google AI Ultra 5x | $100 |
| Google AI Ultra 20x | $200 |
Google’s 2026 subscription update introduced Ultra 5x and reduced the top tier to $200. It also described compute-based five-hour and weekly limits. The multiplier refers to the relevant plan allowance, not a guaranteed number of messages.
The Plus figure is explicitly the published US launch price. Google’s later student-offer terms cite $4.99 or a local equivalent for qualifying non-US Plus renewals and $19.99 for US Pro renewals. These are different markets and offers; use the current Google One plan selector for your account’s exact price and model access.
Google’s credit guide explains that limits and eligible credit use depend on the product. Check the credit activity screen for balance, expiry and supported features. Do not treat an older Flow credit allowance as a guaranteed Gemini chat allowance.
The following are Gemini Developer API Standard text input/output rates, separate from Google One subscriptions:
| Model | Input / output per 1M tokens |
|---|---|
| Gemini 4 Argon | $2 / $10 through 30 Nov 2026; $4 / $20 from 1 Dec |
| Gemini 3.8 Flash / 3.6 Flash | $0.75 / $3.75 through 31 Dec 2026; $1.50 / $7.50 from 1 Jan 2027 |
| Gemini 3.5 Flash-Lite | $0.30 / $2.50 |
| Gemini 3.1 Flash-Lite | $0.25 / $1.50 |
| Gemini 3.1 Pro Preview, including Custom Tools variant | $2 / $12 up to 200K input; $4 / $18 above |
| Gemini 3 Flash Preview | $0.50 / $3 |
| Gemini 2.5 Pro | $1.25 / $10 up to 200K input; $2.50 / $15 above |
| Gemini 2.5 Flash | $0.30 / $2.50 |
| Gemini 2.5 Flash-Lite | $0.10 / $0.40 |
Source: Google’s full API rate card. Audio, images, Live, Omni/video, speech, embeddings and robotics have distinct pricing there. Cached input, storage and grounding can add or change costs; the table covers text rates only.
3. Claude: plan limits plus optional paid continuation
| Plan | Monthly reference price |
|---|---|
| Free | $0 |
| Pro | $20; $200/year with annual billing |
| Max 5x | $100 |
| Max 20x | $200 |
| Team Standard | $25/seat monthly; $20/seat/month annually |
| Team Premium | $125/seat monthly; $100/seat/month annually |
| Enterprise | $20/seat/month annually, plus usage at API rates |
See Claude’s pricing page and individual plan comparison. The Max multipliers describe capacity relative to Pro, not fixed message counts.
Claude’s usage-limit explanation describes five-hour sessions and weekly limits. Model, conversation length, features and effort affect consumption. Optional usage credits continue eligible work beyond included limits at API rates.
Here is a compact view of the models listed on Anthropic’s API rate card. A listed historical rate does not guarantee that a model remains available to your account.
| Claude model | Base input / output per 1M tokens |
|---|---|
| Fable 5.1 / Fable 5 / Mythos 5.1 / Mythos 5 | $10 / $50 |
| Opus 5.5 | $4 / $20 |
| Opus 5 / 4.8 / 4.7 / 4.6 / 4.5 | $5 / $25 |
| Opus 4.1 / 4, historical | $15 / $75 |
| Sonnet 5.5 / 5 | $2 / $10 |
| Sonnet 4.6 / 4.5 / 4 | $3 / $15 |
| Haiku 5.5, up to 100K input | $0.10 / $0.50 |
| Haiku 5.5, above 100K input | $0.50 / $2.50 |
| Haiku 4.5 | $1 / $5 |
| Haiku 3.5, historical | $0.80 / $4 |
Caching, fast mode, residency and tool charges require the full rate card. Some models require special access. Claude app usage credits and any separately granted Console/API credits are different balances; check API billing guidance before assuming one pays for the other.
4. GitHub Copilot: subscription plus a credit allowance
| Individual plan | Monthly price | Monthly AI credits: base + variable flex |
|---|---|---|
| Free | $0 | Limited allowance; Auto selection |
| Pro | $10 | 1,000 + 500 = 1,500 |
| Pro+ | $39 | 3,900 + 3,100 = 7,000 |
| Max | $100 | 10,000 + 10,000 = 20,000 |
Source: GitHub’s individual billing guide. Flex amounts can change. Included credits reset monthly and do not roll over. Paid code completions and next-edit suggestions do not consume AI credits; chat and agent work do. Organisation plans use separate pooled allowances.
One GitHub AI credit = $0.01. A metered model cost of $0.25 consumes 25 credits. That may draw from your allowance rather than create an extra charge; additional paid usage depends on your budget settings.
Copilot default rates: USD per 1M tokens. Grouped base rates match; caching and availability can differ.
| Copilot model | Input / output |
|---|---|
| GPT-5 mini | $0.25 / $2 |
| GPT-5.3-Codex | $1.75 / $14 |
| GPT-5.4 | $2.50 / $15 |
| GPT-5.4 mini | $0.75 / $4.50 |
| GPT-5.4 nano | $0.20 / $1.25 |
| GPT-5.5 | $5 / $30 |
| GPT-5.6 Luna | $0.20 / $1.20 |
| GPT-5.6 Sol | $4 / $20 |
| GPT-5.6 Terra | $2 / $12 |
| GPT-6 Astra | $10 / $50 |
| GPT-6 Luna | $0.10 / $0.50 |
| GPT-6 Sol / 6.1 Sol | $2 / $10 |
| Claude Haiku 4.5 | $1 / $5 |
| Claude Haiku 5.5 | $0.10 / $0.50 |
| Claude Sonnet 4 / 4.6 | $3 / $15 |
| Claude Sonnet 5 / 5.5 | $2 / $10 |
| Claude Opus 4.8 / 5 | $5 / $25 |
| Claude Opus 5.5 | $4 / $20 |
| Claude Opus 4.8 fast preview | $10 / $50 |
| Claude Fable 5 / 5.1 | $10 / $50 |
| Gemini 3.7 Flash / 3.8 Flash | $0.75 / $3.75 |
| Gemini 4 Argon | $2 / $10 |
| MAI-Code-1.1-Flash | $0.20 / $1.20 |
| Grok 4.5 / 4.6 / 4.7 | $2 / $6 |
| Kimi K3 | $3 / $15 |
Check the linked card for long-context, caching and review infrastructure charges.
A small cost example
For 10,000 uncached input tokens and 2,000 output tokens, with no cache writes, tools or long-context premium:
| Direct API model | Calculation | Estimated token cost |
|---|---|---|
| GPT-6.1 Sol | 0.01 × $2 + 0.002 × $10 | $0.04 |
| Gemini 4 Argon, current promotional rate | 0.01 × $2 + 0.002 × $10 | $0.04 |
| Claude Sonnet 5.5 | 0.01 × $2 + 0.002 × $10 | $0.04 |
The same token quantities and base rates in Copilot equal 4 GitHub credits. In the Work/Codex credit table, GPT-6.1 Sol uses 1 credit for those quantities. These are separate billing units; the different credit numbers do not establish a dollar saving. Real tasks can use different token counts across models.
Choose using your own workload
Start with the lowest plan that meets your actual needs. Record how often a limit interrupts work, which model you used and whether a more expensive model improved the result enough to matter. Before enabling extra usage, set an appropriate spending limit and check how that control works.
Use the subscription price to plan a recurring budget. Use the rate card and measured token usage for metered work. Keep those two calculations separate, and revisit the official sources when a plan, model or promotional period changes.