Best LLM Cost Tracking Tools (2026): A Vendor-Neutral Comparison

ArtificialAnalysis is an independent benchmark site tracking 300+ models across 20+ providers — OpenAI, Anthropic, Google, Meta, Mistral, DeepSeek, and more. It's become the de facto pricing reference in the LLM space: its Intelligence Index vs. output-speed charts and cost-per-task figures get cited in vendor launch materials and enterprise benchmark leaderboards alike.

Unlike the gateway and observability tools below, ArtificialAnalysis doesn't touch your production traffic. It answers a different question — "which model should I use" rather than "what did I spend" — and its coverage of effective pricing (cached, batch, and long-context tiers) makes it a genuinely useful cross-check against provider list prices before you commit to an architecture.

Cost-relevant highlights: Cost per million input/output tokens, cost-per-task and tokens-per-task scoring, cached-input and cache-write pricing, updated roughly 8×/day. Free to browse; a paid API (Self-serve/Pro/Commercial tiers) exists for teams that want to pull the data programmatically or redistribute it.

Pricing: The site is free. API access has a free self-serve key for internal use, a paid Pro tier for full data and higher limits, and a custom Commercial tier for redistribution.

Who should use it: Anyone choosing between models before writing code, or periodically re-checking whether a cheaper model now meets the same quality bar. Pair with the effective-cost math in our LLM effective cost table once you know which providers you're comparing.

Watch out for: It's a reference site, not a tracking tool — it tells you nothing about your own production spend. Pricing across providers uses different tokenizers, so a raw $/MTok comparison can understate a model that tokenizes your specific content less efficiently.

LiteLLM is the most widely deployed open-source LLM gateway. It normalizes 100+ models — OpenAI, Anthropic, Gemini, Mistral, Bedrock, Azure, and more — to a single OpenAI-compatible API, meaning you change one line of code and gain routing flexibility across the entire LLM landscape.

For cost tracking specifically, LiteLLM offers: per-model cost calculation using a built-in pricing table (updated regularly), per-user and per-team budget enforcement with hard limits and soft alerts, detailed spend analytics at the model and user level, and SQLite/PostgreSQL storage for historical cost data.

Cost tracking highlights: Set max_budget per user key. Track spend via the /spend/users and /spend/models endpoints. Integrates with Prometheus for infra-level cost dashboards. Logs to Langfuse, Helicone, or Datadog if you want trace-level detail alongside the proxy metrics.

Pricing: Open-source and free to self-host (MIT-licensed Proxy Server). Enterprise tier (self-hosted, license only) starts around $250/month for the basic tier, with reported higher tiers for SLAs, SSO, and audit logs — LiteLLM doesn't publish a clean enterprise pricing table, so confirm current numbers directly with the vendor.

Who should use it: Multi-model teams who need routing + fallbacks + budget enforcement as a unified layer. Excellent choice as the gateway leg of a gateway + observability stack.

Watch out for: Self-hosting adds ops burden. The analytics UI is functional but less polished than dedicated observability tools. LiteLLM disclosed 6 CVEs between March and June 2026 — including a supply-chain compromise and a privilege-escalation chain letting a low-privilege user reach proxy_admin — all since patched; if you self-host an internet-exposed instance, stay current on releases and review the project's security advisories. For trace-level debugging, pair with Langfuse or Braintrust.

Acquired by Mintlify in March 2026 and now in maintenance mode — Helicone's team joined Mintlify, and while security updates, new model support, and bug fixes continue, active new-feature development has slowed. The product remains live and usable at its existing pricing.

Helicone is an analytics-first LLM gateway. The integration is famously simple: add one line changing your base URL (or a single header for non-proxy setups), and every LLM call is logged automatically with cost, latency, token counts, and the full prompt/response.

Unlike LiteLLM, which is primarily a routing layer, Helicone leads with analytics. Its dashboard surfaces cost trends by model, time, user, and property (custom metadata you attach to requests). You can filter to any segment — "show me cost for user_id=123 on Claude-3-Opus over the last 30 days" — without any custom instrumentation beyond adding a metadata header.

Cost tracking highlights: Automatic cost calculation across all major providers. Custom properties for per-user, per-feature attribution. Rate limiting and budget alerts. Prompt management with A/B cost comparison. Open-source (Apache 2.0), self-hostable at every tier.

Pricing: Hobby (free): 10,000 requests/month, 7-day retention. Pro: $79/month, unlimited seats, 1-month retention. Team: $799/month, adds SOC 2 & HIPAA compliance, 3-month retention. Enterprise: custom, on-prem available.

Who should use it: Teams who want detailed LLM analytics with minimal integration work and are comfortable adopting a product in maintenance mode. Strong choice for product teams iterating fast — the prompt playground and cost comparison features make it easy to evaluate prompt changes before deploying.

Watch out for: Routing and fallback features are lighter than LiteLLM. Post-acquisition, expect stability over new features — evaluate whether that trade-off matters for your roadmap.

Acquired by Palo Alto Networks, completed May 29, 2026 — Portkey's AI gateway is now integrated into Palo Alto's Prisma AIRS platform as a unified control plane for governing and securing autonomous AI agents. The core gateway product remains available; expect its roadmap to increasingly reflect Palo Alto's enterprise-security priorities rather than operating as an independent startup.

Portkey is a production AI gateway that combines cost tracking with reliability engineering: automatic fallbacks, load balancing across providers, semantic caching, and canary deployments for prompt changes. For teams where availability and cost are coupled concerns, Portkey addresses both in one layer.

Cost tracking in Portkey is built around "virtual keys" — per-user, per-app API key wrappers that carry budget limits, model access controls, and spend tracking. You create a virtual key for each user or team, set a maximum spend, and Portkey hard-blocks requests when the limit is hit. Cost attribution is automatic — every request is tagged to its virtual key.

Cost tracking highlights: Per-user virtual keys with hard budget limits. Cost breakdowns by model, virtual key, and time window. Semantic caching to avoid re-billing identical prompts (cache hit = $0 cost). Real-time spend dashboards. Audit logs for every request.

Pricing: Developer (free): 10,000 logs/month. Production: $49/month for 100,000 logs. Enterprise: custom. Treat these as pre-acquisition figures — pricing and packaging are explicitly unsettled post-acquisition, so confirm current tiers with Portkey/Palo Alto directly before budgeting.

Who should use it: Production teams who need reliability (fallbacks, load balancing) alongside cost tracking. Particularly useful for multi-tenant applications where each customer needs an isolated budget — and teams already standardizing on Palo Alto's security stack may find the Prisma AIRS integration a plus rather than a concern.

Watch out for: Semantic caching requires careful prompt normalization to hit reliably. The feature set overlaps heavily with LiteLLM and OpenRouter — evaluate all three if you're choosing between gateway tools, and factor in that Portkey's product direction now answers to Palo Alto Networks' roadmap.

OpenRouter is a unified API and model marketplace: one OpenAI-compatible key gets you access to 400+ models from 60+ providers, with automatic routing to the most efficient provider by latency, cost, or quality, plus provider-level failover when a specific host is down. It raised a $113M Series B led by CapitalG (Alphabet's growth fund) in May 2026, more than doubling its valuation to roughly $1.3B in a year — the best-funded tool in this comparison by a wide margin.

Unlike LiteLLM, Helicone, or Portkey, OpenRouter isn't primarily something you self-host or point at your own provider accounts — it's a marketplace you route traffic through, and it charges for that access rather than for observability seats.

Cost-relevant highlights: Cost/latency/quality-based auto-routing ("Auto Exacto"), opt-in model fallback priority lists, "zero-completion insurance" (failed requests aren't billed), response caching, and configurable cost/data-governance guardrails.

Pricing: Passes through provider token rates without markup on standard usage. Charges a 5.5% fee on credit-card top-ups (minimum $0.80), and a 5% fee on bring-your-own-key requests once you exceed 1M/month (5M/month on enterprise plans).

Who should use it: Teams that want the widest possible model selection — including lower-cost open-weight models (DeepSeek, Qwen, GLM) — without integrating each provider separately, and who are comfortable with a marketplace fee model instead of a flat subscription.

Watch out for: The top-up and BYOK fees are a real cost at scale — model them against a direct-provider or self-hosted-gateway approach before committing. It's a marketplace, not an observability tool; pair with Layer 2 for trace-level attribution.

Langfuse is the most widely adopted open-source LLM observability platform. It provides trace-level visibility into every LLM call: input tokens, output tokens, cost (calculated from its model pricing table), latency, user metadata, and the full prompt/completion chain. For multi-step agents and chains, it renders a visual timeline showing how cost and latency decompose across sub-steps.

Cost tracking is first-class in Langfuse. You can view cost by model, by user, by session, or by any custom tag you attach at trace time. The cost dashboard shows daily/weekly/monthly trends with model-level breakdowns. If a new prompt version costs 30% more, you'll see it in the cost trend before it shows up on your API invoice.

Cost tracking highlights: Automatic cost estimation for all major models (you can override with custom pricing). Cost aggregation by user, session, or any custom tag. Prompt management with cost comparison across versions. Evaluation metrics co-located with cost data. Python/TypeScript SDK + OpenAI drop-in wrapper + LangChain integration.

Pricing: Self-hosted: free (MIT license, runs on Docker). Cloud: free tier (50,000 observations/month), Hobby ($59/month), Pro ($119/month), Team ($299/month). Enterprise: custom with SSO and SLA.

Who should use it: The default choice for teams that want trace-level cost visibility without vendor lock-in. Self-hosting on your own infrastructure means no per-observation cost as you scale, and you own the data. Strong choice for any team that isn't already standardized on Datadog or LangSmith.

Watch out for: Self-hosting adds ops burden. The built-in evals are lighter than Braintrust. Langfuse doesn't enforce real-time budget limits — pair with LiteLLM or Portkey for that.

LangSmith is LangChain's observability and evaluation platform. If you're using LangChain or LangGraph, it's the lowest-friction observability option by far — set two environment variables and every chain, agent, and LLM call is automatically traced with cost and latency data. There's no code change required beyond the env vars.

The LangSmith trace UI is purpose-built for LangChain's abstractions: it renders chains, retrievals, tool calls, and agent steps in a nested timeline that mirrors how LangChain actually executes them. Cost is summed at each level of the hierarchy, making it easy to identify which part of a chain is expensive.

Cost tracking highlights: Automatic tracing with zero code changes for LangChain users. Cost aggregated at chain, step, and run level. Dataset-based regression testing with cost comparison across prompt versions. Filter and search runs by model, cost, latency, or custom metadata.

Pricing: Developer (free): limited trace storage. Plus: $39/user/month. Enterprise: custom. Hosted Cloud uses $0.0036/minute standby pricing for LangGraph Cloud deployments — this can accumulate if you forget to scale down.

Who should use it: Teams using LangChain or LangGraph who want zero-friction observability. If you're not on LangChain, Langfuse or Braintrust will give you more for less money.

Watch out for: LangSmith is optimized for the LangChain abstraction layer — it's less useful if you're making raw OpenAI/Anthropic calls or using another framework. Per-user pricing scales poorly for large engineering teams. The LangGraph Cloud standby pricing is easy to forget.

Braintrust is an AI engineering platform that combines evaluation, logging, and prompt management in one product. Its positioning is slightly different from pure-observability tools: it's built for teams running continuous evals as part of their development cycle, not just monitoring in production.

Cost tracking in Braintrust comes alongside evaluation scores — you can see, for any prompt version, both the quality metrics and the cost. This is the right framing for teams optimizing the cost/quality trade-off: you're not just asking "how much did this cost?" but "how much did this cost, and was it worth it?" The experiment tracking UI makes it easy to compare prompt versions on both dimensions simultaneously.

Cost tracking highlights: Cost visible at trace and experiment level. Automatic cost calculation for all major providers. Cost comparison across prompt versions in experiment view. LLM playground with cost preview before deployment. Prompt catalog for managing versions across teams.

Pricing: Braintrust uses usage-based pricing (per log/trace event). Free tier available. Contact for enterprise pricing. The model is friendly for getting started but can scale less predictably at very high log volumes.

Who should use it: Teams who care as much about quality as cost and want to track both in the same tool. Strong choice for ML engineers running structured evals. Less necessary if you're primarily doing production monitoring without a structured eval workflow.

Watch out for: The eval focus can feel like overhead for teams that just want cost tracking. At high log volumes, usage-based pricing requires active monitoring. Less LangChain-native than LangSmith.

Datadog added LLM Observability (GA 2025) as part of its APM platform. For teams already on Datadog, this is the path of least resistance: LLM cost data flows into the same dashboards, alerts, and on-call runbooks as your existing infrastructure metrics. When an LLM cost spike happens at 2am, your on-call engineer already knows how to use the tool that's paging them.

The integration captures token counts, cost estimates, latency, error rates, and custom evaluation metrics for OpenAI, Anthropic, Cohere, and other providers. Auto-instrumentation covers the most common Python and JS clients — no SDK integration required for the basics.

Cost tracking highlights: Auto-instrumentation for OpenAI, Anthropic, Cohere, and LangChain. Cost and latency SLOs in the same Datadog interface as infra SLOs. Anomaly detection on spend using Datadog's existing ML alerting. Trace correlation — link an LLM cost spike to the specific backend service or deployment that caused it.

Pricing: Datadog is the most expensive option on this list. LLM Observability pricing is usage-based on top of existing Datadog contracts. Typical all-in cost for meaningful LLM monitoring is $500–$2,000+/month at scale. Suitable for enterprises already in the Datadog ecosystem where the incremental cost is low relative to existing spend.

Who should use it: Enterprises already standardized on Datadog where adding a new vendor is a harder sell than an incremental Datadog feature. The observability depth for LLM-specific concerns is shallower than dedicated tools, but the platform integration is unmatched.

Watch out for: Expensive to adopt fresh if you're not already a Datadog customer. LLM-specific features (prompt management, eval workflows) are less mature than Langfuse or Braintrust. Not a good choice for cost-sensitive startups.

CloudZero is an established cloud cost intelligence platform ($119M raised across 7 rounds) that launched a dedicated AI financial control plane in May 2026. It ingests real-time AI spend directly from LLM gateways (including LiteLLM), desktop usage, and OpenTelemetry streams, then attributes it by model, provider, prompt pattern, customer, product, and feature — the "unit economics" view that a gateway's per-key budget alone doesn't give you.

This is a finance-facing tool, not a developer-facing one. It doesn't compete with the gateway layer — it consumes the data gateways produce and turns it into cost-per-customer and cost-per-feature numbers a CFO can use.

Cost-relevant highlights: Multi-dimensional AI spend allocation, direct integration with OpenAI, Anthropic, AWS Bedrock, GCP Vertex AI, and Azure OpenAI alongside traditional cloud services, and a conversational "Ask Advisor" assistant for natural-language cost queries.

Pricing: Enterprise, priced as a percentage of managed cloud spend — roughly 1% at $1M/year, declining to 0.6–0.7% at $10M/year. No self-serve tier.

Who should use it: Finance-led organizations with LLM spend well above the ~$50K/month threshold who need cross-cloud, cross-team chargeback reporting — not engineering teams looking for a lightweight tracking tool.

Watch out for: Enterprise sales motion and pricing; overkill below the FinOps threshold. It's a consumer of gateway data, not a replacement for one — you still need Layer 1 or Layer 2 instrumentation feeding it.

I haven't built anything yet — I just need to pick a model

→ Start with ArtificialAnalysis. It's free, covers 300+ models, and its effective-pricing figures (cached/batch included) beat comparing raw list prices across provider pricing pages by hand.

I don't know which model or feature is driving my bill

→ Start with Helicone (lowest integration friction, though now in maintenance mode post-acquisition) or Langfuse (open-source, self-hostable). Both give you per-model, per-user cost attribution within an afternoon of setup. If you're on LangChain, use LangSmith instead — zero code changes.

I need to enforce per-user or per-tenant budget limits

→ Use a gateway tool: LiteLLM for multi-model routing + budget enforcement, Portkey if you also need fallbacks and caching, OpenRouter if you want the widest model marketplace and can absorb its top-up/BYOK fees, Helicone if you want analytics-first with lighter routing. Observability tools alone can't enforce hard limits — they only tell you after the fact.

I'm already on Datadog and don't want a new vendor

→ Enable Datadog LLM Observability. The LLM-specific features are shallower than dedicated tools, but the platform integration with your existing dashboards and alerts is worth the trade-off if Datadog is already your SOT.

I want to compare prompt versions on cost and quality simultaneously

→ Braintrust or LangSmith. Both show cost alongside evaluation metrics in experiment views. Braintrust is the stronger choice for teams with a structured eval workflow; LangSmith for LangChain users.

Finance needs cross-team chargeback reporting, not just engineering visibility

→ CloudZero if you're at real enterprise scale (its pricing is a percentage of managed spend); Costbase or LLM CFO if you're smaller and want a lighter-weight or performance-priced option first.