Pricing Comparison: Anthropic vs Google Gemini pricing - a comparison for AI product builders

Claude and Gemini are the two default alternatives to OpenAI, and their rate cards have moved in opposite directions this year. Gemini is cheaper at every tier on paper, and Anthropic's tokenizer makes the real gap bigger than the AI token pricing tables show.

Anthropic vs Google at a glance

Takeaway: Gemini wins on price at every tier, roughly 2x at the flagship and 7x at the value end, before the Claude tokenizer widens it further.

Model overview

Anthropic sells four tiers. Claude Fable 5, launched in July 2026, is a new premium tier at $10 / $50 with 1M context at standard rates. Claude Opus 4.8 is the flagship at $5 / $25. Claude Sonnet 5 is the mid pick at an introductory $2 / $10 that rises to $3 / $15 on September 1, 2026. Claude Haiku 4.5 is the value tier at $1 / $5.

Google's flagship is Gemini 3.1 Pro, in preview at $2 / $12 for prompts up to 200k. Gemini 3.5 Flash, launched May 19, 2026, is the newest speed tier at $1.50 / $9. Gemini 3.1 Flash-Lite at $0.125 / $0.75 is the value pick, and Gemini 2.5 Pro remains on the card as a legacy model being displaced.

Key features

Pricing snapshot

Prices verified July 13, 2026. Providers reprice often, see the update timeline below.

Anthropic API pricing

Per 1M tokens.

Batch API: 50% off. Cache reads: 0.1x input price. Web search: $10 per 1k searches.

Google API pricing

Per 1M tokens, paid tier, prompts up to 200k where tiered.

Batch: 50% off. Grounded search: $14 per 1k queries after 5,000 free per month. A free tier exists for most Flash models.

Chat plans

What this costs at a real workload

The standardized scenario: 1,000 requests per day at roughly 1,000 input and 1,000 output tokens each, which works out to about 30M input and 30M output tokens per month. Monthly cost = 30 x input rate + 30 x output rate.

Tokenizer caveat for the Claude rows: Fable 5, Sonnet 5, and Opus 4.7+ use a new tokenizer that produces roughly 30% more tokens for the same text. Comparing per-token rates across providers understates Anthropic's effective cost per request by about that margin. Rate cards don't tell you what the invoice will say. The tokenizer does.

Gemini 3.1 Pro at $420 per month runs less than half of Opus 4.8's $900, and about a quarter of Claude Fable 5's $1,800. At the value end the gap is starker: Haiku 4.5 costs $180 against Flash-Lite's roughly $26 on the same math.

The spread across all six providers is 143x between the most and least expensive model for the same workload. The full 15-row table is on the AI Pricing Index. For how per-token rates turn into unit economics, see token economics.

Timeline of past updates

Anthropic

• July 2026: Claude Fable 5 launch at $10 / $50, a new tier priced above Opus

• July 2026: Opus 4.7 fast mode ($30 / $150) deprecated, Opus 4.8 fast mode at $10 / $50

• June 30, 2026: Sonnet 5 launch with a new tokenizer that produces roughly 30% more tokens per text, an effective price change that never appears on the rate card

• June 30, 2026: Sonnet 5 introductory pricing announced at $2 / $10 through August 31, 2026, then $3 / $15. A pre-announced price increase, rare in this market

• November 2025: Opus 4.5 launched at $5 / $25, down from Opus 4.1's $15 / $75. A 3x cut

• October 2025: Haiku 4.5 at $1 / $5, up from Haiku 3.5's $0.80 / $4

Trajectory: Mixed. A big flagship cut in 2025, then upward drift via the premium tier, fast mode, the tokenizer, and the scheduled Sonnet increase.

Google

• May 2026 (I/O): AI Ultra restructured, $99.99 5x tier added, 20x tier cut from $249.99 to $199.99

• May 19, 2026: Gemini 3.5 Flash at $1.50 / $9. 3x over Gemini 3 Flash

• February 19, 2026: Gemini 3.1 Pro preview at $2 / $12, 3.1 Flash-Lite at $0.125 / $0.75

• December 17, 2025: Gemini 3 Flash preview at $0.50 / $3, up 67% from 2.5 Flash

• November 18, 2025: Gemini 3 Pro preview at $2 / $12, up from 2.5 Pro's $1.25 / $10

• June 2025: Gemini 2.5 Flash GA at $0.30 / $2.50

Trajectory: The Flash tier is 5x more expensive than 13 months ago. The era of the ever-cheaper budget model ended here first.

The AI Pricing Index

Both providers logged 6 price events in 12 months. Google's moves concentrated in the Flash tier, up 5x in 13 months. Anthropic cut its flagship 3x in 2025 and has been drifting back up since, through the premium tier, fast mode, the tokenizer, and a scheduled Sonnet increase. Building on either means repricing your own unit economics about every two months.

The index counts documented list-price events per provider over the trailing 12 months, recounted monthly. Full methodology and all six provider timelines are on the AI Pricing Index.

What this means for your own pricing

If you build on Anthropic, Google, or both, this pricing complexity becomes your pricing complexity. Four problems show up regardless of which side you pick.

• The margin problem. Every user interaction has variable cost. A power user generating long responses on a flagship model costs 10-50x more than a casual user on a value model. Per-seat pricing doesn't see this. You need usage-aware billing to protect margins.

• The model mix problem. Products route different requests to different models, and the table above shows the cost range that creates inside a single product. Your billing system needs to know which model served which request and price accordingly.

• The credit translation problem. Many AI products abstract provider costs with credit-based pricing. That means maintaining a conversion layer from usage to credits to dollars. Every provider price event forces you to re-derive the conversion rates.

• The visibility problem. Finance needs margin by customer, by feature, by model, by month. If your metering system lives apart from your billing system, that reconciliation happens in spreadsheets.

The index above adds the repricing angle: these two providers logged 12 price events in 12 months between them. Each one forces you to re-derive your own margins, credit conversion rates, and FX exposure. Solvimon runs that layer: metering per provider, credits and wallets as first-class primitives, rate cards you update without an engineering sprint. See Solvimon for AI.

Related

• OpenAI vs Anthropic pricing. Two flagships priced within dollars, and a tokenizer that changes the math.

• Gemini vs DeepSeek pricing. The two price-aggressive providers, compared on the same workload.