What does an estimated price range mean?

Pages per $1 shows roughly how many pages a dollar buys: higher is better value. Complex pages fall toward the lower end. The cost per page appears underneath. Approximate ranges include extra room for complex pages. Measured ranges start at the 10th percentile and extend to at least twice the 90th percentile, or the highest observed page cost plus 25%, whichever is higher. Benchmark estimates use varied published test documents at current token rates. Typical estimates use at least 30 settled customer requests across 7 days and 3 customers, within the last 90 days and since the last token-rate change. Where measurements are not yet sufficient, we show a broad estimated range of half to twenty times the reference page cost, expanded further when observed costs require it. This planning allowance is not a measured percentile range. Actual charges can fall outside a range; it is not a cap.

Credits are USD. When you add $25 to your account, your balance shows $25.00. Fixed-price engines deduct their per-page charge. Token-billed models deduct the cost of the tokens used, so document complexity and output length affect the final charge. If an engine costs $0.001/page and you process 100 pages, $0.10 is deducted.

Am I charged when a request fails?

It depends on whether the engine ran. If your request never reached an engine — invalid input, a bad API key, a rate limit, an unavailable engine, or any failure on our side — you are not charged. If the engine ran and then failed, you are charged the provider's cost for that attempt, passed through at cost with no markup and never more than the price you were quoted; calling that provider directly would have billed you the same way. Where we have no record of a provider charge for the attempt, there is no charge. Full policy.

No. Your credit balance never expires and can be used with any engine at any time.

What payment methods do you accept?

We accept all major credit and debit cards (Visa, Mastercard, Amex) via Stripe.

Auto top-up automatically adds credits when your balance drops below a threshold you set. For example, you can configure it to add $25 whenever your balance falls below $5. This ensures uninterrupted service.

Do new accounts get starter credits?

Yes. New accounts receive $1.00 in starter credits — no credit card required.

Are there rate limits?

Starter accounts have a 50 requests/day rate limit. After depositing $10 or more, the rate limit increases to 1,000 requests/day.

Do you offer volume discounts?

For high volume needs (millions of pages/month), contact us for custom arrangements.

OpenOCR

OCR pricing

Fixed prices & token estimates$1 free to startNo subscription required

Engine pricing

Compare pages per $1 across all engines; higher means more pages for your money. Per-page charges appear underneath. Token-billed models, marked *, show an estimate for a standard page of about 1,800 characters; denser pages cost more and sparser pages less. Price sorting uses that estimate. What counts as a page, and how are costs estimated?

174 available engines ·

	Provider			Speed
ScaleDown: SummarizeOpenRouter	OOpenRouter🔒 No training	~28,900 pages / $1~$0.0000346 per standard page	No hard score
ScaleDown: ExtractOpenRouter	OOpenRouter🔒 No training	~28,900 pages / $1~$0.0000346 per standard page	No hard score
Z.ai: GLM Flash LatestOpenRouter	OOpenRouter🔒 No training	~12,300 pages / $1~$0.0000815 per standard page	No hard score
Nex AGI: Nex-N2.5-MiniOpenRouter	OOpenRouter🔒 No training	~11,600 pages / $1~$0.0000864 per standard page	No hard score
Qwen: Qwen3.7 FlashOpenRouter	OOpenRouter🔒 No training	~7,910 pages / $1~$0.000126 per standard page	No hard score	slow
Mistral SmallMistral AI	Mistral AI🔒 No training	~4,560 pages / $1~$0.000219 per standard page	53.6%bench	medium
Nex AGI: Nex-N2.5-ProOpenRouter	OOpenRouter🔒 No training	~4,410 pages / $1~$0.000227 per standard page	No hard score
Amazon: Nova Lite 1.0OpenRouter	OOpenRouter🔒 No training	~4,260 pages / $1~$0.000235 per standard page	40.1%bench	medium
ByteDance: UI-TARS 7BOpenRouter	OOpenRouter🔒 No training	~4,180 pages / $1~$0.000239 per standard page	21.5%bench	medium
Gemini 2.5 Flash-LiteGoogle Cloud	Google Cloud🔒 No training	~3,710 pages / $1~$0.00027 per standard page	No hard score	medium
inclusionAI: Ling 3.0 Flash VLOpenRouter	OOpenRouter🔒 No training	~3,330 pages / $1~$0.000301 per standard page	No hard score
Xiaomi: MiMo-V2.6-FlashOpenRouter	OOpenRouter🔒 No training	~3,310 pages / $1~$0.000302 per standard page	No hard score
NVIDIA: Nemotron 3.5 Content SafetyOpenRouter	OOpenRouter🔒 No training	~3,230 pages / $1~$0.000309 per standard page	No hard score
OpenAI: GPT-4.1 NanoOpenRouter	OOpenRouter🔒 No training	~3,040 pages / $1~$0.000329 per standard page	43.5%bench	medium
Xiaomi: MiMo-V2.5OpenRouter	OOpenRouter🔒 No training	~2,620 pages / $1~$0.000381 per standard page	57.3%bench	slow
DeepSeek: DeepSeek Flash LatestOpenRouter	OOpenRouter🔒 No training	~2,480 pages / $1~$0.000404 per standard page	No hard score
Qwen3-VL 30B A3B Instruct (DeepInfra)DeepInfra	DeepInfra🔒 No training	~2,470 pages / $1~$0.000404 per standard page	56.1%bench	medium
OpenAI GPT Luna LatestOpenRouter	OOpenRouter🔒 No training	~2,440 pages / $1~$0.00041 per standard page	No hard score
PrismML: Ternary Bonsai 2 27BOpenRouter	OOpenRouter🔒 No training	~2,240 pages / $1~$0.000447 per standard page	No hard score
Qwen: Qwen3.8 FlashOpenRouter	OOpenRouter🔒 No training	~2,200 pages / $1~$0.000454 per standard page	No hard score
Anthropic: Claude Haiku 5.5OpenRouter	OOpenRouter🔒 No training	~2,200 pages / $1~$0.000455 per standard page	No hard score
Qwen: Qwen3.8 Omni FlashOpenRouter	OOpenRouter🔒 No training	~2,050 pages / $1~$0.000487 per standard page	No hard score
Qwen: Qwen3 VL 30B A3B InstructOpenRouter	OOpenRouter🔒 No training	~1,940 pages / $1~$0.000517 per standard page	56.1%bench	medium
Meta: Llama 4 MaverickOpenRouter	OOpenRouter🔒 No training	~1,560 pages / $1~$0.00064 per standard page	71%bench	medium
Mistral: Mistral Small 3.1 24BOpenRouter	OOpenRouter🔒 No training	~1,510 pages / $1~$0.000663 per standard page	No hard score	medium
Z.ai: GLM 5.3 FlashOpenRouter	OOpenRouter🔒 No training	~1,420 pages / $1~$0.000703 per standard page	No hard score
Meta: Llama 4 ScoutOpenRouter	OOpenRouter🔒 No training	~1,370 pages / $1~$0.000731 per standard page	53.6%bench	medium
DeepSeek: DeepSeek V4 Flash Vision ExpOpenRouter	OOpenRouter🔒 No training	~1,270 pages / $1~$0.000789 per standard page	No hard score
Qwen: Qwen3.6 FlashOpenRouter	OOpenRouter🔒 No training	~1,270 pages / $1~$0.000789 per standard page	49%bench	slow
Qwen: Qwen3.6 35B A3BOpenRouter	OOpenRouter🔒 No training	~1,170 pages / $1~$0.000852 per standard page	43.8%bench	slow
Llama 3.2 11B Vision (DeepInfra)DeepInfra	DeepInfra🔒 No training	~1,170 pages / $1$0.000856/page	No hard score	slow
StepFun: Step 3.7 FlashOpenRouter	OOpenRouter🔒 No training	~1,150 pages / $1~$0.000867 per standard page	63.6%bench	slow
OpenAI: GPT-6 LunaOpenRouter	OOpenRouter🔒 No training	~1,130 pages / $1~$0.000881 per standard page	No hard score
Xiaomi: MiMo-V2.6-ProOpenRouter	OOpenRouter🔒 No training	~1,060 pages / $1~$0.00094 per standard page	No hard score
Qwen3-VL 235B A22B Instruct (DeepInfra)DeepInfra	DeepInfra🔒 No training	~1,000 pages / $1~$0.000996 per standard page	63%bench	slow
Perceptron: Perceptron Mk1OpenRouter	OOpenRouter🔒 No training	~982 pages / $1~$0.00102 per standard page	41.2%bench	slow
Perplexity: SonarOpenRouter	OOpenRouter🔒 No training	~966 pages / $1~$0.00104 per standard page	32.8%bench	medium
MiniMax: MiniMax M3OpenRouter	OOpenRouter🔒 No training	~940 pages / $1~$0.00106 per standard page	55%bench	medium
Perceptron: Perceptron Mk1.5OpenRouter	OOpenRouter🔒 No training	~895 pages / $1~$0.00112 per standard page	No hard score
DeepSeek: DeepSeek V4.1 FlashOpenRouter	OOpenRouter🔒 No training	~852 pages / $1~$0.00117 per standard page	No hard score
Mistral: Mistral Large 3 2512OpenRouter	OOpenRouter🔒 No training	~849 pages / $1~$0.00118 per standard page	58.5%bench	medium
Gemini 3.1 Flash Lite PreviewGoogle Cloud	Google Cloud🔒 No training	~845 pages / $1~$0.00118 per standard page	60.5%bench	fast
Gemini 3.1 Flash LiteGoogle Cloud	Google Cloud🔒 No training	~821 pages / $1~$0.00122 per standard page	60.7%bench	fast
OpenAI: GPT-5.6 LunaOpenRouter	OOpenRouter🔒 No training	~816 pages / $1~$0.00123 per standard page	64.8%bench	medium
Cohere: Command A+OpenRouter	OOpenRouter🔒 No training	~812 pages / $1~$0.00123 per standard page	No hard score
Google: Gemini 3.1 Flash LiteOpenRouter	OOpenRouter🔒 No training	~784 pages / $1~$0.00128 per standard page	60.7%bench	medium
Meta: Muse Glimmer 30BOpenRouter	OOpenRouter🔒 No training	~754 pages / $1~$0.00133 per standard page	49.2%bench	slow
Mistral MediumMistral AI	Mistral AI🔒 No training	~739 pages / $1~$0.00135 per standard page	62.1%bench	medium
Qwen: Qwen3.7 PlusOpenRouter	OOpenRouter🔒 No training	~731 pages / $1~$0.00137 per standard page	No hard score	slow
Mistral: Mistral Medium 3OpenRouter	OOpenRouter🔒 No training	~684 pages / $1~$0.00146 per standard page	64.9%bench	medium
Mistral: Mistral Medium 3.1OpenRouter	OOpenRouter🔒 No training	~684 pages / $1~$0.00146 per standard page	64.6%bench	medium
Qwen: Qwen3 VL 235B A22B InstructOpenRouter	OOpenRouter🔒 No training	~678 pages / $1~$0.00148 per standard page	63%bench	medium
AWS TextractAWS	AWS🔒 No training	~623 pages / $1$0.001605/page	19.5%bench	medium
Mistral MediumMistral AI	Mistral AI🔒 No training	~602 pages / $1~$0.00166 per standard page	61.9%bench	medium
Amazon: Nova 2 LiteOpenRouter	OOpenRouter🔒 No training	~580 pages / $1~$0.00172 per standard page	41.6%bench	medium
Z.ai: GLM 5.3 FlashXOpenRouter	OOpenRouter🔒 No training	~571 pages / $1~$0.00175 per standard page	No hard score
Mistral Large 4Mistral AI	Mistral AI🔒 No training	~556 pages / $1~$0.0018 per standard page	No hard score
Gemini 3.5 Flash LiteGoogle Cloud	Google Cloud🔒 No training	~544 pages / $1~$0.00184 per standard page	56.9%bench	fast
Mistral: Mistral Large 4OpenRouter	OOpenRouter🔒 No training	~515 pages / $1~$0.00194 per standard page	No hard score
Google: Gemini 3.5 Flash-LiteOpenRouter	OOpenRouter🔒 No training	~504 pages / $1~$0.00198 per standard page	56.9%bench	medium
EasyOCROpen-OCR.com	Open-OCR.com🔒 No training	500 pages / $1$0.002/page	4.8%bench	medium
Tesseract 5Open-OCR.com	Open-OCR.com🔒 No training	500 pages / $1$0.002/page	No hard score	medium
OCRmyPDFOpen-OCR.com	Open-OCR.com🔒 No training	500 pages / $1$0.002/page	No hard score	medium
docTROpen-OCR.com	Open-OCR.com🔒 No training	500 pages / $1$0.002/page	12.8%bench	medium
RapidOCROpen-OCR.com	Open-OCR.com🔒 No training	500 pages / $1$0.002/page	No hard score	fast
Mistral LargeMistral AI	Mistral AI🔒 No training	~477 pages / $1~$0.0021 per standard page	58.5%bench	medium
Qwen2.5-VL 32B Instruct (DeepInfra)DeepInfra	DeepInfra🔒 No training	~467 pages / $1$0.00214/page	51.6%bench	slow
Qwen: Qwen3.8 27BOpenRouter	OOpenRouter🔒 No training	~455 pages / $1~$0.0022 per standard page	47.3%bench	slow
Qwen: Qwen3.5 Plus 2026-04-20OpenRouter	OOpenRouter🔒 No training	~415 pages / $1~$0.00241 per standard page	No hard score	slow
GPT 4.1 MiniOpenAI	OpenAI🔒 No training	~412 pages / $1~$0.00243 per standard page	46.2%bench	medium
GPT 4o MiniOpenAI	OpenAI🔒 No training	~401 pages / $1~$0.00249 per standard page	32.3%bench	medium
OpenAI: GPT-6 Luna ProOpenRouter	OOpenRouter🔒 No training	~393 pages / $1~$0.00254 per standard page	No hard score
Thinking Machines: Inkling SmallOpenRouter	OOpenRouter🔒 No training	~389 pages / $1~$0.00257 per standard page	43.4%bench	medium
StepFun: Step 5 PreviewOpenRouter	OOpenRouter🔒 No training	~383 pages / $1~$0.00261 per standard page	No hard score
Qwen: Qwen3.6 27BOpenRouter	OOpenRouter🔒 No training	~365 pages / $1~$0.00274 per standard page	45.1%bench	slow
Pareto 26.10 PreviewOpenRouter	OOpenRouter🔒 No training	~362 pages / $1~$0.00276 per standard page	No hard score
Gemini 3.8 FlashGoogle Cloud	Google Cloud🔒 No training	~351 pages / $1~$0.00285 per standard page	No hard score
Google: Gemini 3.8 FlashOpenRouter	OOpenRouter🔒 No training	~325 pages / $1~$0.00308 per standard page	No hard score
ByteDance Seed: Seed 2.1 TurboOpenRouter	OOpenRouter🔒 No training	~320 pages / $1~$0.00313 per standard page	No hard score	slow
Amazon: Nova Pro 1.0OpenRouter	OOpenRouter🔒 No training	~320 pages / $1~$0.00313 per standard page	21.1%bench	medium
xAI: Grok 4.3OpenRouter	OOpenRouter🔒 No training	~318 pages / $1~$0.00314 per standard page	54.3%bench	medium
OpenAI GPT Mini LatestOpenRouter	OOpenRouter🔒 No training	~313 pages / $1~$0.00319 per standard page	No hard score	medium
Claude Haiku 4.5Anthropic	Anthropic🔒 No training	~300 pages / $1~$0.00333 per standard page	42.6%bench	medium
MoonshotAI: Kimi K2.7 CodeOpenRouter	OOpenRouter🔒 No training	~288 pages / $1~$0.00347 per standard page	No hard score	slow
Anthropic: Claude Haiku 4.5OpenRouter	OOpenRouter🔒 No training	~274 pages / $1~$0.00365 per standard page	42.6%bench	medium
Meta: Muse Spark 1.3OpenRouter	OOpenRouter🔒 No training	~261 pages / $1~$0.00383 per standard page	No hard score
ByteDance Seed: Seed-2.0-CodeOpenRouter	OOpenRouter🔒 No training	~259 pages / $1~$0.00385 per standard page	53%bench	slow
IBM Docling (hosted)IBM Docling (hosted)	IIBM Docling (hosted)🔒 No training	250 pages / $1$0.004/page	No hard score
Thinking Machines: InklingOpenRouter	OOpenRouter🔒 No training	~240 pages / $1~$0.00416 per standard page	No hard score	medium
Anthropic Claude Haiku LatestOpenRouter	OOpenRouter🔒 No training	~231 pages / $1~$0.00433 per standard page	No hard score	medium
MoonshotAI: Kimi K2.6OpenRouter	OOpenRouter🔒 No training	~230 pages / $1~$0.00436 per standard page	No hard score	slow
Google: Gemini 3.7 FlashOpenRouter	OOpenRouter🔒 No training	~226 pages / $1~$0.00442 per standard page	51.6%bench	medium
Gemini 3.7 FlashGoogle Cloud	Google Cloud🔒 No training	~217 pages / $1~$0.0046 per standard page	51.6%bench	medium
Mistral OCR 4.1Mistral AI	Mistral AI🔒 No training	200 pages / $1$0.005/page	No hard score
Google: Gemini 3.6 FlashOpenRouter	OOpenRouter🔒 No training	~189 pages / $1~$0.00529 per standard page	56.5%bench	medium
OpenAI: GPT-5.6 Luna ProOpenRouter	OOpenRouter🔒 No training	~189 pages / $1~$0.0053 per standard page	53.9%bench	slow
Gemini 3 Flash PreviewGoogle Cloud	Google Cloud🔒 No training	~184 pages / $1~$0.00543 per standard page	63.6%bench	slow
Mistral: Mistral Medium 3.5OpenRouter	OOpenRouter🔒 No training	~182 pages / $1~$0.0055 per standard page	No hard score	medium
Gemini 3.6 FlashGoogle Cloud	Google Cloud🔒 No training	~179 pages / $1~$0.00559 per standard page	56.5%bench	medium
SpaceXAI: Grok 4.7OpenRouter	OOpenRouter🔒 No training	~178 pages / $1~$0.00562 per standard page	No hard score
Sakana: Fugu MaxOpenRouter	OOpenRouter🔒 No training	~178 pages / $1~$0.00562 per standard page	No hard score
Qwen: Qwen3.8 Max (0902)OpenRouter	OOpenRouter🔒 No training	~178 pages / $1~$0.00562 per standard page	No hard score
xAI: Grok 4.5OpenRouter	OOpenRouter🔒 No training	~176 pages / $1~$0.00567 per standard page	56.5%bench	slow
Nano Banana 2.1Google Cloud	Google Cloud🔒 No training	~175 pages / $1~$0.0057 per standard page	No hard score
GPT 4.1OpenAI	OpenAI🔒 No training	~174 pages / $1~$0.00575 per standard page	36.1%bench	medium
Google Gemini Flash LatestOpenRouter	OOpenRouter🔒 No training	~166 pages / $1~$0.00601 per standard page	No hard score	medium
Mistral Medium 3 5Mistral AI	Mistral AI🔒 No training	~166 pages / $1~$0.00603 per standard page	No hard score	medium
Google: Nano Banana 2.1OpenRouter	OOpenRouter🔒 No training	~162 pages / $1~$0.00616 per standard page	No hard score
Qwen: Qwen3.8 MaxOpenRouter	OOpenRouter🔒 No training	~157 pages / $1~$0.00636 per standard page	No hard score	slow
Meta: Muse Spark 1.1OpenRouter	OOpenRouter🔒 No training	~143 pages / $1~$0.007 per standard page	No hard score	medium
Pareto 26.9OpenRouter	OOpenRouter🔒 No training	~142 pages / $1~$0.00702 per standard page	No hard score
Pareto LatestOpenRouter	OOpenRouter🔒 No training	~142 pages / $1~$0.00702 per standard page	No hard score
OpenAI: GPT-5.6 SolOpenRouter	OOpenRouter🔒 No training	~139 pages / $1~$0.00721 per standard page	36.1%bench	slow
Anthropic Claude Sonnet LatestOpenRouter	OOpenRouter🔒 No training	~136 pages / $1~$0.00733 per standard page	No hard score	slow
Anthropic: Claude Sonnet 5OpenRouter	OOpenRouter🔒 No training	~136 pages / $1~$0.00733 per standard page	47.4%bench	medium
OpenAI: GPT-5.1-CodexOpenRouter	OOpenRouter🔒 No training	~135 pages / $1~$0.0074 per standard page	33.5%bench	slow
Gemini 4 ArgonGoogle Cloud	Google Cloud🔒 No training	~132 pages / $1~$0.0076 per standard page	No hard score
Google: Gemini 4 ArgonOpenRouter	OOpenRouter🔒 No training	~122 pages / $1~$0.00821 per standard page	No hard score
Anthropic: Claude Sonnet 5.5OpenRouter	OOpenRouter🔒 No training	~122 pages / $1~$0.00821 per standard page	No hard score
OpenAI: GPT-6 Sol ProOpenRouter	OOpenRouter🔒 No training	~122 pages / $1~$0.00821 per standard page	No hard score
OpenAI: GPT-6.1 SolOpenRouter	OOpenRouter🔒 No training	~122 pages / $1~$0.00821 per standard page	No hard score
OpenAI GPT Sol LatestOpenRouter	OOpenRouter🔒 No training	~122 pages / $1~$0.00821 per standard page	No hard score
OpenAI: GPT-6 SolOpenRouter	OOpenRouter🔒 No training	~122 pages / $1~$0.00821 per standard page	No hard score
OpenAI: GPT-6.1 Sol ProOpenRouter	OOpenRouter🔒 No training	~122 pages / $1~$0.00821 per standard page	No hard score
OpenAI: GPT-5.6 TerraOpenRouter	OOpenRouter🔒 No training	~118 pages / $1~$0.00851 per standard page	29.5%bench	medium
Xiaomi: MiMo-V2.6-Pro-UltraSpeedOpenRouter	OOpenRouter🔒 No training	~106 pages / $1~$0.0094 per standard page	No hard score
OpenAI GPT Terra LatestOpenRouter	OOpenRouter🔒 No training	~105 pages / $1~$0.0095 per standard page	No hard score
Meta: Muse Spark 1.2OpenRouter	OOpenRouter🔒 No training	~103 pages / $1~$0.00975 per standard page	65.5%bench	slow
Datalab Convert (accurate)Datalab	DDatalab🔒 No training	100 pages / $1$0.01/page	77.2%bench	slow
SpaceXAI: Grok 4.6OpenRouter	OOpenRouter🔒 No training	~99.4 pages / $1~$0.0101 per standard page	41.1%bench	medium
Perplexity: Sonar ProOpenRouter	OOpenRouter🔒 No training	~91.9 pages / $1~$0.0109 per standard page	37.1%bench	medium
xAI: Grok Build 0.1OpenRouter	OOpenRouter🔒 No training	~89.5 pages / $1~$0.0112 per standard page	45.1%bench	slow
Qwen: Qwen3.8 Max PrimeOpenRouter	OOpenRouter🔒 No training	~89 pages / $1~$0.0112 per standard page	No hard score
Amazon: Nova Premier 1.0OpenRouter	OOpenRouter🔒 No training	~87 pages / $1~$0.0115 per standard page	No hard score	slow
Fireworks: Ember-1OpenRouter	OOpenRouter🔒 No training	~81.2 pages / $1~$0.0123 per standard page	No hard score
OpenAI: GPT-5.6 Sol ProOpenRouter	OOpenRouter🔒 No training	~77.3 pages / $1~$0.0129 per standard page	33.7%bench	slow
Anthropic: Claude Sonnet 4OpenRouter	OOpenRouter🔒 No training	~76.9 pages / $1~$0.013 per standard page	49.5%bench	medium
Anthropic: Claude Sonnet 4.5OpenRouter	OOpenRouter🔒 No training	~76.9 pages / $1~$0.013 per standard page	50.3%bench	slow
Anthropic: Claude Sonnet 4.6OpenRouter	OOpenRouter🔒 No training	~76.9 pages / $1~$0.013 per standard page	54.1%bench	slow
MoonshotAI: Kimi K3OpenRouter	OOpenRouter🔒 No training	~75.2 pages / $1~$0.0133 per standard page	No hard score	slow
O4 MiniOpenAI	OpenAI🔒 No training	~72.1 pages / $1~$0.0139 per standard page	41.7%bench	slow
MoonshotAI Kimi LatestOpenRouter	OOpenRouter🔒 No training	~68.5 pages / $1~$0.0146 per standard page	No hard score	slow
Anthropic: Claude Opus LatestOpenRouter	OOpenRouter🔒 No training	~68.2 pages / $1~$0.0147 per standard page	No hard score	slow
Meta: Muse WonderOpenRouter	OOpenRouter🔒 No training	~60.9 pages / $1~$0.0164 per standard page	No hard score
Anthropic: Claude Opus 5.5OpenRouter	OOpenRouter🔒 No training	~60.9 pages / $1~$0.0164 per standard page	No hard score
Google: Gemini 3.5 FlashOpenRouter	OOpenRouter🔒 No training	~59.6 pages / $1~$0.0168 per standard page	44.2%bench	slow
Claude Opus 4.5Anthropic	Anthropic🔒 No training	~59.2 pages / $1~$0.0169 per standard page	63.1%bench	slow
o3OpenAI	OpenAI🔒 No training	~54.7 pages / $1~$0.0183 per standard page	30.4%bench	slow
Anthropic: Claude Opus 4.8OpenRouter	OOpenRouter🔒 No training	~54.6 pages / $1~$0.0183 per standard page	No hard score	slow
Gemini 3.5 FlashGoogle Cloud	Google Cloud🔒 No training	~51.3 pages / $1~$0.0195 per standard page	44.2%bench	medium
OpenAI: GPT-5.6 Terra ProOpenRouter	OOpenRouter🔒 No training	~50.7 pages / $1~$0.0197 per standard page	37.6%bench	slow
OpenAI: GPT-5.4 Image 2OpenRouter	OOpenRouter🔒 No training	~48.2 pages / $1~$0.0207 per standard page	31.6%bench	medium
OpenAI: GPT Chat LatestOpenRouter	OOpenRouter🔒 No training	~47.6 pages / $1~$0.021 per standard page	No hard score	medium
Anthropic: Claude Opus 4.5OpenRouter	OOpenRouter🔒 No training	~46.1 pages / $1~$0.0217 per standard page	67.5%bench	slow
Anthropic: Claude Opus 4.6OpenRouter	OOpenRouter🔒 No training	~46.1 pages / $1~$0.0217 per standard page	63.4%bench	slow
Claude Opus 5OpenRouter	OOpenRouter🔒 No training	~45.2 pages / $1~$0.0221 per standard page	45.6%bench	slow
Sakana: Fugu Ultra v2OpenRouter	OOpenRouter🔒 No training	~42.1 pages / $1~$0.0238 per standard page	No hard score
Google: Gemini 2.5 Pro Preview 06-05OpenRouter	OOpenRouter🔒 No training	~41.3 pages / $1~$0.0242 per standard page	25.6%bench	slow
OpenAI: GPT-5.5OpenRouter	OOpenRouter🔒 No training	~35.6 pages / $1~$0.0281 per standard page	29.2%bench	slow
Google Gemini Pro LatestOpenRouter	OOpenRouter🔒 No training	~27.6 pages / $1~$0.0362 per standard page	No hard score	slow
Anthropic: Claude Fable LatestOpenRouter	OOpenRouter🔒 No training	~27.3 pages / $1~$0.0367 per standard page	No hard score	slow
Anthropic: Claude Fable 5OpenRouter	OOpenRouter🔒 No training	~27.3 pages / $1~$0.0367 per standard page	50.8%bench	slow
Gemini 2.5 ProGoogle Cloud	Google Cloud🔒 No training	~25.1 pages / $1~$0.0398 per standard page	25.6%bench	slow
Anthropic: Claude Fable 5.1OpenRouter	OOpenRouter🔒 No training	~24.4 pages / $1~$0.041 per standard page	No hard score
OpenAI: GPT-6 AstraOpenRouter	OOpenRouter🔒 No training	~24.4 pages / $1~$0.041 per standard page	No hard score
OpenAI: GPT-6 Astra ProOpenRouter	OOpenRouter🔒 No training	~24.4 pages / $1~$0.041 per standard page	No hard score
OpenAI GPT Astra LatestOpenRouter	OOpenRouter🔒 No training	~24.4 pages / $1~$0.041 per standard page	No hard score
Kimi K3 (Fireworks)Fireworks AI	Fireworks AI🔒 No training	~24.1 pages / $1~$0.0414 per standard page	No hard score
Gemini 3.1 Pro PreviewGoogle Cloud	Google Cloud🔒 No training	~21.7 pages / $1~$0.046 per standard page	30.4%bench	slow
Gemini 3.1 Pro Preview Custom ToolsGoogle Cloud	Google Cloud🔒 No training	~18.3 pages / $1~$0.0546 per standard page	30.4%bench	slow
Anthropic: Claude Opus 4.1OpenRouter	OOpenRouter🔒 No training	~18.3 pages / $1~$0.0547 per standard page	No hard score	slow
Sakana: Fugu UltraOpenRouter	OOpenRouter🔒 No training	~13.4 pages / $1~$0.0748 per standard page	No hard score	slow
o1OpenAI	OpenAI🔒 No training	~10.8 pages / $1~$0.0922 per standard page	No hard score	slow
OpenAI: GPT-5.5 ProOpenRouter	OOpenRouter🔒 No training	~7.88 pages / $1~$0.127 per standard page	No hard score