Claude 4 Sonnet API Rate Limits & Pricing 2026

AnthropicGenerally available

API model string

claude-sonnet-5

Context window

1M tokens

Max output 128K tokens (sync), 300K tokens (batch)

Knowledge cutoff: not published
Released: June 30, 2026
Modalities: text + image in, text out

Last verified July 10, 2026

Rate limits by tier

Anthropic replaced its numeric Tier 1–4 system with Start, Build, Scale (+ Custom) on June 26, 2026. Monthly spend caps: Start $500, Build $1,000, Scale $200,000. As of June 2026, Sonnet 4.x/5 shares the same RPM/ITPM/OTPM ceilings as Opus 4.x at every tier. Sonnet 4.x/5 models share one org-level pool. Note: Sonnet 5 is NOT supported on Priority Tier per Anthropic service-tiers docs (July 2026). Figures are per Anthropic docs via third-party trackers — verify actual limits in your console at platform.claude.com/settings/limits.

Tier	Requirements	RPM	TPM	RPD	Concurrent	Notes
Start	First API key or cumulative deposit under $40	~50 (per third-party trackers — confirm in your console/dashboard)	~30,000 ITPM (legacy Tier-1 equivalent; anecdotal post-Colossus reports ~500,000 ITPM — per third-party trackers, confirm in your console/dashboard)	not published	not published	Testing and prototyping only. Sonnet 5 NOT on Priority Tier. Spend cap $500/month.
Build	Cumulative spend above ~$40; automatic promotion	not published	not published	not published	not published	Automatic on crossing threshold. RPM and ITPM scale above Start. Spend cap $1,000/month.
Scale	Cumulative deposit threshold (~$200 equivalent); spend cap $200,000/month	~4,000 (per third-party trackers — confirm in your console/dashboard)	~2,000,000 ITPM; ~400,000–800,000 OTPM (per third-party trackers — confirm in your console/dashboard)	not published	not published	Primary production planning ceiling. Cache reads excluded from ITPM for Sonnet 5. Higher limits via Custom tier. Sonnet 5 NOT on Priority Tier regardless of tier level.
Custom	Contact Anthropic sales	not published	not published	not published	not published	No spend cap. Negotiated RPM/ITPM/OTPM. Dedicated support SLA. Priority Tier still not available for Sonnet 5 at Custom tier.

Swipe the table sideways to see every limit column.

1.Token-bucket replenishment is continuous. Acceleration limits can trigger 429 on sudden spikes even within tier ceilings — ramp gradually, add sleep between request bursts.
2.Cache reads do NOT count toward ITPM for Sonnet 5 (current non-† model). This is the key throughput optimization for high-context Sonnet workloads.
3.Sonnet 5 Priority Tier restriction is per-model, not per-tier. This applies even at Custom tier. If Priority Tier is required, use Claude Opus 4.8.

Limits verified against the Anthropic docs, July 10, 2026.

Token pricing

What you pay per million tokens (USD). Input and output are billed separately.

Input

$2.00

per 1M tokens

Output

$10.00

per 1M tokens

Cached input

$0.20

per 1M cached

Batch discount

−50%

on batch jobs

INTRODUCTORY pricing through August 31, 2026. Standard pricing flips to $3.00 input / $15.00 output per MTok on September 1, 2026. High urgency: batch and cache-heavy workloads should be scheduled before this date.
Cache read = $0.20/MTok (0.1× base input $2.00). 5-minute cache write = $2.50/MTok (1.25× base). 1-hour cache write = $4.00/MTok (2× base). Cache reads excluded from ITPM.
Batch API (50% off): introductory $1.00 input / $5.00 output per MTok. Async within 24 hours. Max 300K output tokens per request with output-300k-2026-03-24 beta header.
Extended thinking tokens billed as output (can multiply output cost 3–5×). Disable for non-reasoning tasks.
US-only data residency (inference_geo: 'us') adds 1.1× surcharge on Sonnet 4.6 and later. Omit unless legally required.
Web search tool adds $10 per 1,000 searches on top of token costs. Cache tool-calling contexts to minimize repeat searches.
No long-context surcharge on 1M-context Sonnet 5. No free API tier. No first-party embeddings or fine-tuning.

Side-project chatbot

$20

per month

Assumptions

5M input tokens / 1M output tokens per month, no caching

5M × $2.00/MTok = $10 input + 1M × $10.00/MTok = $10 output = $20/mo

Mid-scale app with caching

$164.40

per month

Assumptions

60M input / 12M output per month, 70% of input served from cache

18M × $2.00/MTok = $36.00 (non-cached) + 42M × $0.20/MTok = $8.40 (cache reads) + 12M × $10.00/MTok = $120.00 output = $164.40/mo

High-volume production

$1,600

per month

Assumptions

400M input / 80M output per month, no caching

400M × $2.00/MTok = $800 input + 80M × $10.00/MTok = $800 output = $1,600/mo

Run your own numbers

Drag your real monthly token volumes and watch the bill update live — priced against rival models at the same usage.

Input tokens / month30M

1M500M

Output tokens / month5M

0.1M100M

Estimated claude-sonnet-5 spend

$110/mo

Input: $60

Output: $50

30M in × $2.00 + 5M out × $10.00 = $110

Same volume, priced across models

Gemini 2.5 ProCheapest$88
claude-sonnet-5This model$110
Claude 4 Opus$275
GPT-5$300

Rivals priced at their published input/output rates for the same monthly volumes. Prompt caching is model-specific, so it is applied to claude-sonnet-5 only. Estimates for comparison; real bills vary with request shape and long-context surcharges.

claude-sonnet-5 vs the alternatives

Claude Sonnet 5 competes with Claude Opus 4.8 (higher capability, higher cost), GPT-5.5 (OpenAI), and Gemini 2.5 Pro (Google) on context, pricing, and latency.

Aspect	claude-sonnet-5	Claude 4 Opus	GPT-5	Gemini 2.5 Pro
Context window	1M tokens	1M tokens	~1M tokens (922K input)	1M+ tokens
Input $/MTok (intro)	$2.00	$5.00	$5.00	$1.25 (≤200K)
Output $/MTok (intro)	$10.00	$25.00	$30.00	$10.00
Cached input $/MTok	$0.20	$0.50	$0.50	~$0.125
Batch discount	50%	50%	50%	50%
Long-context surcharge	none	none	2× above 272K	2× above 200K (full prompt)
Priority Tier support	NOT supported	supported	supported	n/a
Free API tier	no	no	no	paid-only since Apr 1, 2026
Max output (sync)	128K	128K	128K	not published

Swipe the table sideways to see every model.

Hitting a 429? The playbook

The exact errors you'll see

rate_limit_errortype: "rate_limit_error"HTTP 429 Too Many Requests

Why it happens & how to fix it

RPM burst: sending requests too fast within a short window

Token-bucket + acceleration limits fire on sudden bursts even within tier ceilings. Ramp gradually; add sleep between request bursts. Smooth request curves over minutes, not seconds.

ITPM saturation on high-context inputs or repeated system prompts

Prompt caching is the key fix: cache reads are excluded from ITPM for Sonnet 5. Use cache_control breakpoints on large system prompts and repeated conversation context. Use 1-hr cache writes ($4.00/MTok) for long-lived contexts.

OTPM exceeded on bulk generation workloads

Reduce max_tokens per request. Move bulk generation to Batch API (50% off, async within 24h, up to 300K output tokens per request) to remove OTPM pressure from the real-time path.

Attempting to route Sonnet 5 to Priority Tier

Sonnet 5 is NOT supported on Priority Tier per Anthropic service-tiers docs (July 2026). This is a per-model restriction. Use standard Start/Build/Scale tier path. If Priority Tier is required, switch to Claude Opus 4.8.

Multi-team org Sonnet pool depleted

Query the Rate Limits API (GET /v1/rate-limits, launched April 25, 2026) to see remaining headroom. Partition API keys by team and monitor anthropic-ratelimit-requests-remaining and anthropic-ratelimit-tokens-remaining on each response.

Retry strategy

On receiving a 429, read the retry-after header (seconds to wait before retrying). Parse anthropic-ratelimit-requests-remaining, anthropic-ratelimit-tokens-remaining, anthropic-ratelimit-reset-tokens, and anthropic-ratelimit-reset-requests to identify whether you are RPM-bound or TPM-bound. If TPM-bound, consider reducing max_tokens on the retry. Use exponential backoff with jitter: start at 1 second, double each retry, cap at 64 seconds. Do not retry immediately.

retry.ts

1const ANTHROPIC_API_URL = "https://api.anthropic.com/v1/messages";
2
3async function callSonnetWithRetry(
4  payload: object,
5  maxRetries = 5
6): Promise<Response> {
7  let attempt = 0;
8  let delayMs = 1000;
9
10  while (attempt <= maxRetries) {
11    const res = await fetch(ANTHROPIC_API_URL, {
12      method: "POST",
13      headers: {
14        "x-api-key": process.env.ANTHROPIC_API_KEY!,
15        "anthropic-version": "2023-06-01",
16        "content-type": "application/json",
17      },
18      body: JSON.stringify(payload),
19    });
20
21    if (res.status !== 429) return res;
22
23    const retryAfter = res.headers.get("retry-after");
24    const reqsRemaining = res.headers.get("anthropic-ratelimit-requests-remaining");
25    const toksRemaining = res.headers.get("anthropic-ratelimit-tokens-remaining");
26    const resetReqs = res.headers.get("anthropic-ratelimit-reset-requests");
27
28    console.warn(
29      `429 on attempt ${attempt + 1}/${maxRetries + 1}. ` +
30      `Requests left: ${reqsRemaining} (reset: ${resetReqs}), ` +
31      `Tokens left: ${toksRemaining}. Retry-After: ${retryAfter}s`
32    );
33
34    const waitMs = retryAfter
35      ? parseInt(retryAfter, 10) * 1000
36      : delayMs + Math.random() * 500;
37
38    await new Promise((r) => setTimeout(r, waitMs));
39    delayMs = Math.min(delayMs * 2, 64000);
40    attempt++;
41  }
42
43  throw new Error(`Exceeded max retries (${maxRetries}) on 429`);
44}

How to raise your limits

The ladder from the starter tier to enterprise — what each rung takes, and what it unlocks.

Start

Immediate

Automatic on creating your first API key or depositing under $40 cumulative. No application required.

Unlocks: ~50 RPM (per third-party trackers). Spend cap $500/month. Testing and prototyping only. Priority Tier not available.

Build

Automatic within hours to days of crossing threshold

Automatic when cumulative spend crosses ~$40. No application required. Monitor at platform.claude.com/settings/limits.

Unlocks: RPM and ITPM ceilings scale above Start. Spend cap $1,000/month. Suitable for early production.

Scale

Automatic within days of crossing threshold

Automatic when cumulative deposit crosses ~$200. Check platform.claude.com/settings/limits for position. Query GET /v1/rate-limits for programmatic ceiling monitoring.

Unlocks: ~4,000 RPM / ~2,000,000 ITPM / ~400,000–800,000 OTPM (per third-party trackers — confirm in console). Spend cap $200,000/month. Cache reads excluded from ITPM. Primary production ceiling.

Custom

Days to weeks to negotiate

Contact Anthropic sales. Discuss volume commitments and annual agreements. Enterprise agreements can accelerate Custom access.

Unlocks: No spend cap. Negotiated RPM/ITPM/OTPM above Scale. Dedicated support SLA. Note: Priority Tier remains unavailable for Sonnet 5 even at Custom — this is a per-model restriction.

Cut your token spend

Prompt caching

−90% cost on repeated context; cache reads excluded from ITPM — dual throughput and cost benefit

Add cache_control breakpoints to system prompts and high-frequency context blocks. Cache reads cost $0.20/MTok (0.1× base) AND do not consume ITPM quota. Use 1-hr cache writes ($4.00/MTok) for long-running sessions. This is the single biggest lever for high-context Sonnet 5 workloads.

Batch API before September 1, 2026

50% off intro pricing = $1.00/$5.00 per MTok; removes OTPM pressure from real-time path

Route all non-latency-sensitive jobs to Batch API now while introductory pricing applies. Async within 24h. Max 300K output tokens per request with output-300k-2026-03-24 beta header. After Sept 1 batch pricing also adjusts.

Lock in workloads before September 1, 2026

Saves 33–50% versus post-September standard pricing

Intro pricing ($2.00/$10.00) ends August 31, 2026. Standard price becomes $3.00/$15.00 on September 1. Schedule high-volume batch runs, evaluation datasets, and bulk jobs before this date.

Ramp traffic gradually

Avoids acceleration-triggered 429s within tier ceilings

Anthropic's acceleration limits are separate from tier ceilings. Smooth request curves over minutes, not seconds. Add queuing or sleep intervals between burst requests.

Disable extended thinking on non-reasoning tasks

Avoids 3–5× output cost multiplier

Extended thinking tokens bill as output. Enable only for tasks genuinely requiring multi-step reasoning. Disable for summarization, extraction, classification, and routing tasks.

Skip Priority Tier routing

Prevents wasted latency and configuration errors

Sonnet 5 is NOT on Priority Tier. Attempting to route to Priority wastes request latency. Use standard tier path; for Priority Tier requirements, switch to Claude Opus 4.8.

Skip US-only data residency unless required

Saves 10% on applicable Sonnet versions

inference_geo: 'us' adds 1.1× surcharge on Sonnet 4.6 and later. Omit unless you have a specific legal or compliance requirement for US data residency.

Cache web search tool results

Reduces $10/1K searches surcharge on repeated queries

Web search tool charges $10 per 1,000 searches on top of tokens. Cache tool-calling contexts containing search results to avoid repeated lookups for the same queries across sessions.

Frequently asked questions

Is claude-sonnet-4 still working?

No. The claude-sonnet-4 API string stopped accepting requests on June 15, 2026. Any calls using that model string will fail. Migrate to claude-sonnet-5, which is Anthropic's current default price/performance model released June 30, 2026.

How do I increase Claude Sonnet 5 rate limits?

Rate limits increase automatically: ~$40 cumulative spend promotes you to Build; ~$200 cumulative unlocks Scale (~4,000 RPM / ~2M ITPM per third-party trackers). Above Scale, contact Anthropic sales for Custom tier with negotiated limits and no spend cap. Check your current tier at platform.claude.com/settings/limits or query GET /v1/rate-limits.

Is the Claude Sonnet 5 API free?

No. There is no free tier. All token usage is billed from the Start tier. Introductory pricing is $2.00 input / $10.00 output per MTok through August 31, 2026, after which standard pricing ($3.00/$15.00) takes effect.

How does Claude Sonnet 5 compare to Gemini 2.5 Pro pricing?

At intro rates, Sonnet 5 is $2.00 input / $10.00 output per MTok vs Gemini 2.5 Pro at $1.25 input (≤200K context) / $10.00 output. Gemini 2.5 Pro has a cheaper base input rate for short contexts, but Sonnet 5 has no long-context surcharge on its 1M window while Gemini charges 2× above 200K tokens. Cache reads: Gemini ~$0.125/MTok vs Sonnet 5 $0.20/MTok. Both offer 50% batch discounts.

What does Claude API rate_limit_error mean?

rate_limit_error (HTTP 429) means you hit RPM, ITPM, or OTPM limits. Check anthropic-ratelimit-requests-remaining to see if it is RPM-bound; check anthropic-ratelimit-tokens-remaining for TPM. Read retry-after header for how many seconds to wait. For Sonnet 5, the most common cause is ITPM saturation from large contexts — activating prompt caching resolves this since cache reads are excluded from ITPM.

Can Claude Sonnet 5 be used on Priority Tier?

No. Per Anthropic's service-tiers documentation (July 2026), Claude Sonnet 5 is NOT supported on Priority Tier. This is a per-model restriction that applies at all tiers including Custom. If Priority Tier access is required for your use case, use Claude Opus 4.8 instead.

When does Claude Sonnet 5 introductory pricing end?

Introductory pricing ($2.00/$10.00 per MTok) ends August 31, 2026. Starting September 1, 2026, standard pricing applies at $3.00 input / $15.00 output per MTok. Batch API pricing adjusts proportionally. This is a high-urgency deadline — schedule bulk workloads before August 31.

How can I reduce Claude API costs for a multi-team organization using Sonnet 5?

Top approaches: (1) Prompt caching — $0.20/MTok cache reads excluded from ITPM; (2) Batch API — $1.00/$5.00 at intro rates (50% off); (3) Partition API keys by team to monitor pool usage; (4) Query GET /v1/rate-limits to track remaining headroom per dimension; (5) Route classification and routing tasks to Haiku 4.5 ($1/$5) and reserve Sonnet 5 for generation. RapidDev helps teams implement multi-model routing and quota engineering — book a free scoping call at rapidevelopers.com/contact.

What is the difference between Claude 4 Sonnet and Claude Sonnet 5?

Claude 4 Sonnet (claude-sonnet-4) was the earlier snapshot in the Sonnet 4.x series that stopped serving on June 15, 2026. Claude Sonnet 5 (claude-sonnet-5), released June 30, 2026, is the current GA model and Anthropic's default recommendation. They are in the same Sonnet 4.x/5 org-level pool and carry the same tier limits. Use claude-sonnet-5 for all new work.

We build AI apps that don't hit rate limits

Retry, backoff & caching built in
Multi-provider fallback routing
Fixed price, you own the code

Get a free estimate

30-min call. No commitment.

Claude 4 Sonnet API Rate Limits, Pricing & Performance (July 2026)

Rate limits by tier

Token pricing

Run your own numbers

claude-sonnet-5 vs the alternatives

Hitting a 429? The playbook

How to raise your limits

Start

Build

Scale

Custom

Cut your token spend

Prompt caching

Batch API before September 1, 2026

Lock in workloads before September 1, 2026

Ramp traffic gradually

Disable extended thinking on non-reasoning tasks

Skip Priority Tier routing

Skip US-only data residency unless required

Cache web search tool results

Frequently asked questions

We build AI apps that don't hit rate limits

Your next step

Still weighing your options?

We put the rapid in RapidDev

Claude 4 Sonnet API Rate Limits, Pricing & Performance (July 2026)

Rate limits by tier

Token pricing

Run your own numbers

claude-sonnet-5 vs the alternatives

Hitting a 429? The playbook

How to raise your limits

Start

Build

Scale

Custom

Cut your token spend

Prompt caching

Batch API before September 1, 2026

Lock in workloads before September 1, 2026

Ramp traffic gradually

Disable extended thinking on non-reasoning tasks

Skip Priority Tier routing

Skip US-only data residency unless required

Cache web search tool results

Related models

Claude 4 Opus API Rate Limits & Pricing

Claude 3.5 Sonnet API Rate Limits & Pricing

Claude 3 Haiku API Rate Limits & Pricing

GPT-5 API Rate Limits & Pricing

More claude-sonnet-5 resources

Claude API 429 & 500 Error Fixes

Claude Sonnet 5 in API Automations

Claude Sonnet 5 for AI Feature Implementation

Frequently asked questions

We build AI apps that don't hit rate limits

Your next step

Still weighing your options?

We put the rapid in RapidDev