How do I track OpenAI API costs?

Install the Cohrint SDK with one line: pip install cohrint. Wrap your OpenAI client with createOpenAIProxy() and every call is automatically tracked with cost, tokens, latency and model details.

Which LLM providers does Cohrint support?

Cohrint supports OpenAI, Anthropic (Claude), Google (Gemini), Mistral, Cohere, Meta (Llama) and any OpenAI-compatible API endpoint.

Yes. Cohrint is free to get started with no credit card required. The dashboard, SDK and API are all free for individual developers and small teams.

How does Cohrint compare LLM pricing?

Cohrint includes a real-time model pricing calculator that compares costs across GPT-4o, Claude 3.5 Sonnet, Gemini 1.5 Pro, Llama 3 and 30+ other models based on your exact token usage.

Cohrint — AI Coding Tool Cost Intelligence Platform

// capabilities

Everything you need to run AI intelligently

📊

Token & cost analytics

40+ metrics

Real-time breakdown of token usage and spend per model, feature, user and team. See which 5% of requests consume 50% of your budget.

→ per-request granularity · daily trend · 365-day history

💱

Cross-model pricing intelligence

24 LLMs live

Live pricing across 24 LLMs from OpenAI, Anthropic, and Google. Compare costs side-by-side and find the cheapest model for your exact workload and quality requirements.

→ "save $3,200/mo switching to Gemini Flash"

🏷️

Cost attribution & chargeback

per-call tags

Break down AI spend by team, feature, user or customer. Tag every call with custom dimensions. Built-in team budgets with per-team alerts and RBAC data isolation.

→ team · feature · user · environment · custom tags

🔔

Budget alerts & enforcement

4 thresholds

Configurable spend thresholds at the team and organizational level. Graduated alerts notify stakeholders progressively as budgets are approached, with automated escalation before limits are breached.

→ graduated alerts · per-team · per-org · Slack · email

📈

Exec ROI dashboard

cost / outcome

Translate token costs into business outcomes — cost per PR, per resolved ticket, per feature shipped. Built for CTOs and CFOs who need signal, not raw token counts.

→ cost/PR · cost/commit · lines-of-code per dollar

♻️

Cache & waste detection

cut redundant spend

Identify redundant AI spend before it compounds. Cohrint surfaces patterns of repeated work across your team's usage, enabling targeted caching decisions that translate directly to cost reduction.

→ redundancy analysis · savings by model · cost trend reporting

👥

Developer productivity ROI

per-developer

Track cost per PR, cost per commit, and lines of code per dollar for every developer. Correlate AI spend with actual code output. Enterprise chargeback reporting built in.

→ cost/PR · cost/commit · lines/$ · team budgets

🚨

Anomaly detection

proactive alerts

Continuous monitoring of your AI spend patterns. Detects cost spikes, traffic surges, and runaway agents — with Slack notifications before the damage reaches your invoice.

→ continuous monitoring · Slack · cost spike detection

⚙️

CI/CD cost gate

pipeline guard

Enforce AI spend policies directly in your CI/CD pipeline. Deployments that exceed defined thresholds are flagged automatically, ensuring budget governance extends into your engineering workflow.

→ GitHub Actions · configurable thresholds · automated enforcement

🤖

MCP — query costs from your IDE

12 tools

Native integration with Claude Code, VS Code, and any MCP-compatible editor. Ask cost questions in natural language and receive actionable insights without leaving your development environment.

→ Claude Code native · VS Code · any MCP editor · zero configuration

⌨️

Cohrint CLI agent

instant visibility

A lightweight terminal companion for AI-assisted development. Tracks spend, surfaces cost insights, and provides real-time feedback — all without interrupting your existing workflow.

→ works with any AI CLI · real-time tracking · zero friction

🔒

Local proxy — zero trust

3 privacy modes

Run the Cohrint proxy on your machine. API keys and prompts never leave your environment — only token counts and costs reach the dashboard. Strict / standard / relaxed modes.

→ strict · standard · relaxed · npm install cohrint-local-proxy

🌐

Cross-platform OTel collector

10+ tools

Unified cost visibility across every AI coding tool your team uses. Plug in once and Cohrint automatically consolidates spend, tokens, and usage data from all connected tools into a single dashboard.

→ Claude Code · GitHub Copilot · Gemini · Codex · and more

🔗

Agent trace explorer

end-to-end visibility

Visualize multi-step AI agent workflows end-to-end. Understand cost and latency distribution across every step in a pipeline — enabling precise optimization at the workflow level, not just the call level.

→ workflow visualization · cost per step · latency breakdown

💡

Live recommendations

actionable

Continuous, automated analysis of your AI usage patterns. Cohrint surfaces prioritized cost-reduction opportunities and notifies your team via Slack before overspend reaches your invoice.

→ model optimization · proactive alerts · Slack notifications

🐍

SDK — Python & TypeScript

2 lines

Drop-in proxy wrappers for OpenAI and Anthropic SDKs. Zero API changes. Streaming supported. Works in any framework. Auto-captures cost, tokens, latency and metadata per call.

→ pip install cohrint · npm install cohrint · streaming

🛡️

Security & data governance

zero plaintext

Enterprise-grade security architecture with credentials stored in hardened, non-reversible form. All session management follows current web security standards. Hosted on Cloudflare's globally distributed, SOC 2-certified edge network.

→ hardened credential storage · secure sessions · SOC 2 · Cloudflare edge

👤

RBAC & team scoping

4 roles

Four roles: owner / admin / member / viewer. Members can be scoped to a single team — they see only their team's data, never another team's. Viewers are read-only at the API level.

→ owner · admin · member · viewer · team-scoped isolation

📋

Audit log & compliance

every action

Full event stream of every API action — who accessed what, when, from which key. Export-ready for compliance audits, security reviews, and SOC2 evidence packages.

→ actor · action · timestamp · exportable

🎯

Hallucination & quality scoring

6 dimensions

Track hallucination rate, faithfulness, relevancy, consistency, toxicity and efficiency per model and per call. Compare quality across providers and catch regressions before users do.

→ hallucination · faithfulness · relevancy · toxicity · efficiency

🔔

Budget alerts & enforcement

4 thresholds

Configurable spend thresholds at the team and organizational level. Graduated alerts notify stakeholders progressively as budgets are approached, with automated escalation before limits are breached.

→ graduated alerts · per-team · per-org · Slack · email

🔑

Key recovery & session security

single-use tokens

Secure, time-limited account recovery delivered via email. Session tokens carry strong entropy and are bound to authenticated devices. Administrators can invalidate all active sessions instantly.

→ time-limited recovery · device-bound sessions · instant revocation

⚙️

CI/CD cost gate

pipeline guard

Enforce AI spend policies directly in your CI/CD pipeline. Deployments that exceed defined thresholds are flagged automatically, ensuring budget governance extends into your engineering workflow.

→ GitHub Actions · configurable thresholds · automated enforcement

// integration

Two lines. Seriously that's it.

Python

TypeScript

MCP (Claude Code)

CLI Wrapper

# Before
from openai import OpenAI

# After — only 2 lines changed
import cohrint
from cohrint.proxy.openai_proxy import OpenAI

cohrint.init(api_key="crt_your_key")
client = OpenAI(api_key="sk-...")

# Everything else is identical — Cohrint wraps transparently
response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "Hello!"}],
)
# ✓ Tokens: 12 in, 8 out  ✓ Cost: $0.000110  ✓ Latency: 423ms
# ✓ Cheapest alternative: gemini-1.5-flash — save 94%
// Before
import OpenAI from "openai";

// After — only 2 lines changed
import { init, createOpenAIProxy } from "cohrint";
import OpenAI from "openai";

init({ apiKey: "crt_your_key" });
const openai = createOpenAIProxy(new OpenAI());

// Identical API — Cohrint wraps every call automatically
const response = await openai.chat.completions.create({
  model: "gpt-4o",
  messages: [{ role: "user", content: "Hello!" }],
});
// ✓ Captured: tokens, cost, latency, cheapest alternative
// ~/.claude/mcp.json  (or claude-code / VS Code equivalent)
{
  "mcpServers": {
    "cohrint": {
      "command": "npx",
      "args": ["-y", "cohrint-mcp"],
      "env": {
        "COHRINT_API_KEY": "crt_your_key",
        "COHRINT_ORG_ID": "your_org_id"
      }
    }
  }
}

// Then ask Claude Code in chat:
// "How much did I spend on AI this week?"
// "Which model is cheapest for my summarisation workflow?"
// "Show requests wasting the most tokens"
# Install globally
$ npx cohrint-cli

# Pipe mode — optimize + forward to Claude
$ echo "Could you please explain kubernetes pods" | cohrint
  ⚡ Optimized: 16 → 12 tokens (saved 4, -25%)

  A Kubernetes pod is the smallest deployable unit...

  💰 Cost: $0.0065 | 💾 Saved: 4 tokens

# REPL mode — switch agents on the fly
$ cohrint
cohrint [claude] ▸ explain load balancers
cohrint [claude] ▸ /gemini summarize this in 2 lines
cohrint [claude] ▸ /compare what is DNS
cohrint [claude] ▸ /session  # interactive mode with /compact, /clear
cohrint [claude] ▸ /summary # dashboard stats in terminal

// live demo

See it in action — interactive preview

cohrint.com / app.html

INTERACTIVE

MTD Spend

$4,821

↑ 12%

Tokens Used

182M

↓ eff +8%

Developers

3 providers

Budget Used

64%

$3,086 left

Daily spend by provider — last 30 days

CLAUDE CODE

$2,140

8 developers

CODEX CLI

$1,430

5 developers

GEMINI CLI

$820

4 developers

Agent	Model	Time	Tokens	Cost
★ gemini	gemini-2.0-flash	1.2s	320	$0.0001
claude	claude-sonnet-4-6	2.1s	420	$0.0065
codex	gpt-4o	1.8s	380	$0.0048

Create free account → Open full dashboard →

No credit card required · Free tier includes 50,000 events/month

Capability	✦ Cohrint	API Gateway Tools	LLM Observability SDKs	General APM Platforms
Real-time cost dashboard	✓	✓	partial	✓
Cross-model pricing intelligence	✓	—	—	—
Token efficiency scoring	✓	—	—	—
MCP server (AI-native IDE integration)	✓	—	—	—
Budget alerts (Slack/email)	✓	✓	partial	✓
Cross-platform OTel ingestion	✓	—	—	—
CLI agent wrapper + optimizer	✓	—	—	—
Privacy-first local proxy	✓	—	—	—
RBAC + full audit log	✓	partial	partial	✓
Free tier	✓ generous	✓	✓	—

// faq

Common questions

Does Cohrint store my prompts and responses? +

Cohrint stores a short preview of your request and response (first 500 chars) to help you debug expensive or low-quality calls. You can disable this entirely in your SDK config. We never train models on your data, and you can request deletion at any time.

Does adding the Cohrint SDK add latency to my calls? +

No measurable latency is added. Cohrint captures event data after your LLM call completes and sends it to our ingest API in a background thread. Your main request path is completely unaffected — zero added latency.

What happens if the Cohrint ingest API is down? +

Your AI calls continue to work normally — Cohrint is not in the critical path. Events are queued locally and retried when connectivity is restored. Your production traffic is never affected by Cohrint availability.

Which LLM providers and models are supported? +

Cohrint supports models across 3 providers: OpenAI (GPT-4o, o1, o3, Codex), Anthropic (Claude 3.5/4 family), and Google (Gemini 1.5/2.0). New models are added within 24 hours of launch.

Can I self-host Cohrint on my own infrastructure? +

Yes. The Cohrint server is open-source and deployable on Railway, Render, or any VPS in under 10 minutes. Enterprise customers get a Docker Compose package and Kubernetes Helm chart for air-gapped deployments where data never leaves your VPC.

How does the free tier compare to paid plans? +

The free tier gives you 50,000 events/month, the core cost dashboard, 7-day log retention, and 1 user seat — no credit card required, forever. The Team plan ($99/mo) adds unlimited requests, cross-model pricing intelligence, budget alerts, team attribution, and up to 10 seats.

What is the Cohrint CLI and how does it work? +

The Cohrint CLI is a lightweight terminal companion that integrates alongside any AI coding agent. It tracks usage costs in real time, provides session-level spend summaries, and surfaces optimization opportunities as you work — without modifying your existing workflow or requiring any configuration changes.

How does cross-platform OTel tracking work? +

Cohrint integrates with Claude Code, Codex CLI, Gemini CLI, and more via standard OpenTelemetry. Setup takes under 2 minutes — no code changes required. Token counts, cost metrics, and session data flow in automatically. For tools that don't report costs natively, Cohrint calculates USD spend from token counts using live model pricing.

Know what your AI bill will be
before it arrives — and cut it.

Stop guessing.
Start measuring.

Know what your AI bill will bebefore it arrives — and cut it.

Stop guessing.Start measuring.

Know what your AI bill will be
before it arrives — and cut it.

Stop guessing.
Start measuring.