PHPackages token27/nexus-ai-pricing - PHPackages - PHPackages [Skip to content](#main-content)[PHPackages](/)[Directory](/)[Categories](/categories)[Trending](/trending)[Leaderboard](/leaderboard)[Changelog](/changelog)[Analyze](/analyze)[Collections](/collections)[Log in](/login)[Sign up](/register) 1. [Directory](/) 2. / 3. [Utility & Helpers](/categories/utility) 4. / 5. token27/nexus-ai-pricing ActiveLibrary[Utility & Helpers](/categories/utility) token27/nexus-ai-pricing ======================== Financial cost calculation for LLM API usage. Supports token-accurate pricing for OpenAI, Anthropic (with Prompt Caching discounts), Google Gemini, DeepSeek, Groq, Mistral, and more. v1.1.0(2mo ago)0492MITPHP ^8.3 Since May 21Compare [ Source](https://github.com/token27/nexus-ai-pricing)[ Packagist](https://packagist.org/packages/token27/nexus-ai-pricing)[ Docs](https://github.com/token27/nexus-ai-pricing)[ RSS](/packages/token27-nexus-ai-pricing/feed)WikiDiscussions Synced 3w ago READMEChangelog (2)Dependencies (4)Versions (3)Used By (2) nexus-ai-pricing ================ [](#nexus-ai-pricing) [![CI](https://github.com/token27/nexus-ai-pricing/actions/workflows/ci.yml/badge.svg)](https://github.com/token27/nexus-ai-pricing/actions)[![PHPStan Level 8](https://camo.githubusercontent.com/412205ac46adf8d1c9329cfcacca7da2697c664b7f1ffd18076a4a17c1e9de6d/68747470733a2f2f696d672e736869656c64732e696f2f62616467652f5048505374616e2d4c6576656c253230382d316636666562)](https://phpstan.org/)[![Latest Version](https://camo.githubusercontent.com/c2213a007b54fbe80a26038731734f6183528786dd61d154bf17c43ae4d4ed8f/68747470733a2f2f696d672e736869656c64732e696f2f7061636b61676973742f762f746f6b656e32372f6e657875732d61692d70726963696e672e7376673f7374796c653d666c61742d737175617265)](https://packagist.org/packages/token27/nexus-ai-pricing)[![PHP 8.3+](https://camo.githubusercontent.com/38027453aeb7eb818641c9de8f82b7624c3558d92634f1946edc715c3ddf8956/68747470733a2f2f696d672e736869656c64732e696f2f62616467652f5048502d382e332532422d3737374242343f6c6f676f3d706870266c6f676f436f6c6f723d7768697465)](https://php.net)[![License: MIT](https://camo.githubusercontent.com/fdf2982b9f5d7489dcf44570e714e3a15fce6253e0cc6b5aa61a075aac2ff71b/68747470733a2f2f696d672e736869656c64732e696f2f62616467652f4c6963656e73652d4d49542d79656c6c6f772e737667)](LICENSE)[![Tests](https://camo.githubusercontent.com/c318016c0fd259debf6db852016128fcbe17a85094056bd0835c2b1ef730302d/68747470733a2f2f696d672e736869656c64732e696f2f62616467652f54657374732d373225323070617373696e672d627269676874677265656e)](docs/testing.md) An **extensively engineered, framework-agnostic** PHP 8.3+ financial calculation library precisely mapped for LLM API computations. Support inherently covers mathematically accurate pricing for OpenAI, Anthropic, Google Gemini, DeepSeek, Groq, Mistral, and Perplexity — resolving everything from Prompt Caching variants to multimodal image tiles locally. Why nexus-ai-pricing? --------------------- [](#why-nexus-ai-pricing) Calculating AI costs is chaotic. OpenAI utilizes a *subset caching* methodology where read tokens are subtracted from prompt tokens. Anthropic relies on *additive cache writes* with massive entry surcharges. Gemini alters limits based on *context tier tiers* and Perplexity adds *per-request bounds*. **nexus-ai-pricing** solves this permanently by: - **Mathematical Accuracy:** Handling all floating-point conversions safely without exception flaws. - **Provider-agnostic core:** Evaluate cache, images, and text outputs using homogeneous DTOs (`PricingResult`). - **Zero-Latency Estimations:** Utilizing [`token27/nexus-ai-tokenizer`](../nexus-ai-tokenizer) heavily to count everything strictly offline before issuing network requests. - **Graceful Degradation:** Safely tags unknown models without triggering fatal system crashes during recursive loops. Features -------- [](#features) - **Universal Support**: Standard metrics evaluated and natively bundled for ChatGPT, Claude, Gemini, DeepSeek Flash, and more. - **Advanced Caching Mathematics**: Separating explicitly standard text costs from cached outputs seamlessly. - **Vision Dimension Math**: Evaluates visual inputs correctly processing parameter rules dynamically (Tiles vs Dimensions). - **Chained Price Tables**: Override built-in arrays securely via `JsonFilePriceTable` structures explicitly resolving priority overrides. - **Glob Pattern Matcher**: Registers dynamic limits using wildcard variables (e.g. `gemini-2.1-*`) safely. - **Type Safety**: PHPStan Level 8, completely production-grade testing limits. Installation ------------ [](#installation) ``` composer require token27/nexus-ai-pricing ``` **Requires:** PHP 8.3+ · [`token27/nexus-ai-tokenizer`](https://packagist.org/packages/token27/nexus-ai-tokenizer) `^1.0` Quick Start ----------- [](#quick-start) ### 1. Post-Request Billing calculation (Using API returned values) [](#1-post-request-billing-calculation-using-api-returned-values) ``` use Token27\NexusAI\Pricing\Engine\PricingEngine; $result = PricingEngine::for('claude-sonnet-4-6')->calculate( inputTokens: 250, outputTokens: 400, cacheWriteTokens: 800, // Anthropic 5-minute surcharge cacheReadTokens: 5000 // 90% discounted block ); echo $result->totalCostUsd(); // Output standard metrics seamlessly echo $result->cacheSavingsUsd(); // Output exactly what was saved against baselines ``` ### 2. Pre-Request Proactive Estimations [](#2-pre-request-proactive-estimations) ``` // Processes offline immediately avoiding Network HTTP limits $estimated = PricingEngine::for('gpt-4o-mini') ->estimate('Generate a dense historical assessment regarding the Roman Empire.'); echo $estimated->format(); ``` ### 3. Multimodal Image Evaluation [](#3-multimodal-image-evaluation) ``` use Token27\NexusAI\Pricing\ValueObject\ImageAttachment; // Handles internal tile mapping explicitly $visionPrice = PricingEngine::for('gpt-4o')->estimateWithImages( text: 'Analyze the architectural defect visible here', images: [ ImageAttachment::highDetail(1920, 1080) ] ); echo $visionPrice->imageCostUsd(); ``` ### 4. Custom Image Estimators [](#4-custom-image-estimators) All three pricing axes are independently injectable. For proprietary or custom vision models, pass your own `ImageTokenEstimatorInterface` implementations: ``` use Token27\NexusAI\Pricing\Engine\PricingEngine; use Token27\NexusAI\Pricing\PriceTable\ArrayPriceTable; use Token27\NexusAI\Pricing\ValueObject\ModelPrice; use Token27\Tokenizer\Contract\ImageTokenEstimatorInterface; use Token27\Tokenizer\Contract\TokenCountInterface; use Token27\Tokenizer\Registry\TokenizerRegistry; use Token27\Tokenizer\ValueObject\TokenCount; use Token27\Tokenizer\Vision\AnthropicImageEstimator; use Token27\Tokenizer\Vision\GeminiImageEstimator; use Token27\Tokenizer\Vision\OpenAIImageEstimator; // Proprietary model: flat 500 tokens per image regardless of size/detail $myEstimator = new class implements ImageTokenEstimatorInterface { public function estimateImageTokens(int $widthPx, int $heightPx, string $detail, string $model): TokenCountInterface { return new TokenCount(count: 500, model: $model, strategy: 'flat_500', approximate: false); } public function supports(string $model): bool { return str_starts_with($model, 'my-vision-'); } }; $engine = new PricingEngine( tokenizer: TokenizerRegistry::createDefault(), priceTable: new ArrayPriceTable([ new ModelPrice('my-vision-model', inputPerMillion: 1.00, outputPerMillion: 4.00, imageInputPerMillion: 2.00), ]), imageEstimators: [ $myEstimator, // my-vision-* → 500 tokens flat (checked first) new OpenAIImageEstimator(), // gpt-4o, o1, o3, … (built-in) new AnthropicImageEstimator(), // claude-* (built-in) new GeminiImageEstimator(), // gemini-* (built-in) ], ); $result = $engine->make('my-vision-model')->estimateWithImages( text: 'Analyse this diagram.', images: [ImageAttachment::highDetail(1920, 1080), ImageAttachment::lowDetail(800, 600)], ); echo $result->imageTokens(); // 1000 (2 images × 500 flat) echo $result->imageCostUsd(); ``` Resolution: the first estimator whose `supports($model)` returns `true` wins. When `imageEstimators` is empty (the default), all three built-in estimators are used automatically. Documentation ------------- [](#documentation) - [Getting Started](docs/getting-started.md) — Cost calculations and quick one-liners - [Installation Setup](docs/installation.md) — Composer, requirements, and DI container setup - [Architecture Schematic](docs/architecture.md) — Core concepts, pricing engine flows and data structures - [Pricing Engine Rules](docs/pricing-engine.md) — Class APIs, estimate methods, and facades - [Pricing Result Typings](docs/pricing-result.md) — Returned metrics, DTOs, and standard formatting - [Caching Strategies](docs/caching-strategies.md) — Anthropic Additive Logic vs OpenAI Subset bounds - [Multimodal Vision Pricing](docs/multimodal-vision.md) — Image complexities, tiles, constraints, custom image estimators - [Custom Price Tables](docs/custom-tables.md) — Decorating interfaces using external JSON arrays - [Pricing Registry](docs/pricing-registry.md) — Factory instantiation, memory safety - [Advanced Integration](docs/advanced-integration.md) — Immutable recursive iteration - [Examples Guide](docs/examples.md) — Directory map against all 8 baseline examples - [Default Internal Catalog](docs/default-catalog.md) — Complete parameter specifications - [System Internals](docs/internals.md) — Math limits and unknown models - [Testing & Validation](docs/testing.md) — PHPUnit, CS-Fixer and logic overrides - [Contributing Guidelines](docs/contributing.md) — Safe repository upgrades License ------- [](#license) MIT. See [LICENSE](LICENSE). ### Health Score 40 — FairBetter than 86% of packages Maintenance87 Actively maintained with recent releases Popularity8 Limited adoption so far Community6 Small or concentrated contributor base Maturity50 Maturing project, gaining track record How is this calculated?**Maintenance (25%)** — Last commit recency, latest release date, and issue-to-star ratio. Uses a 2-year decay window. **Popularity (30%)** — Total and monthly downloads, GitHub stars, and forks. Logarithmic scaling prevents top-heavy scores. **Community (15%)** — Contributors, dependents, forks, watchers, and maintainers. Measures real ecosystem engagement. **Maturity (30%)** — Project age, version count, PHP version support, and release stability. ### Release Activity Cadence Every ~0 days Total 2 Last Release 64d ago ### Community Maintainers ![](https://avatars.githubusercontent.com/u/78189912?v=4)[Token27](/maintainers/token27)[@token27](https://github.com/token27) --- Tags tokensopenaiGeminillmanthropicdeepseekcostpricing ### Code Quality TestsPHPUnit Static AnalysisPHPStan Code StylePHP CS Fixer Type Coverage Yes ### Embed Badge ![Health badge](/badges/token27-nexus-ai-pricing/health.svg) ``` [![Health](https://phpackages.com/badges/token27-nexus-ai-pricing/health.svg)](https://phpackages.com/packages/token27-nexus-ai-pricing) ``` ### Alternatives [cognesy/instructor-php The complete AI toolkit for PHP: unified LLM API, structured outputs, agents, and coding agent control 326123.0k1](/packages/cognesy-instructor-php)[symfony/ai-platform PHP library for interacting with AI platform provider. 521.4M300](/packages/symfony-ai-platform)[llm-agents/agents LLM Agents PHP SDK - Autonomous Language Model Agents for PHP 16716.2k9](/packages/llm-agents-agents)[soukicz/llm LLM client with support for cache, tools and async requests 4414.3k](/packages/soukicz-llm) PHPackages © 2026 [Directory](/)[Categories](/categories)[Trending](/trending)[Changelog](/changelog)[Analyze](/analyze)