CodeBurn – Analyze Claude Code token usage by task
Built this after realizing I was spending ~$1400/week on Claude Code with almost no visibility into what was actually consuming tokens.Tools like ccusage give a cost breakdown per model and per day, but I wanted to understand usage at the task level.CodeBurn reads the JSONL session transcripts that Claude Code stores locally (~/.claude/projects/) and classifies each turn into 13 categories based on tool usage patterns (no LLM calls involved).One surprising result: about 56% of my spend was on conversation turns with no tool usage. Actual coding (edits/writes) was only ~21%.The interface is an interactive terminal UI built with Ink (React for terminals), with gradient bar charts, responsive panels, and keyboard navigation. There’s also a SwiftBar menu bar integration for macOS.Happy to hear feedback or ideas.
FL score
out of 100
Verdict
high confidence
Competition
7
competitors found, emerging market, funded players
Trend
No signal yet
A hyper-niche tool for Claude Code users to analyze task-level token usage from local sessions, revealing hidden costs and enabling optimization.
The pain
The gap
Build angle
Strengths
- Solves a very specific, high-severity financial pain point for a clear target audience.
- Founder has direct experience with the problem and has already built an MVP.
- Unique data source (local Claude Code transcripts) and analysis method provide a strong niche moat.
- Clear, measurable value proposition (cost savings, usage optimization).
- High willingness to pay due to significant potential cost savings.
- Strong potential for viral adoption within niche communities.
- The market for LLM cost optimization is rapidly growing.
Risks
- Relies solely on Claude Code's local session transcript format, which could change, breaking the tool.
- The niche, while valuable, might be smaller than broader LLM observability markets.
- Convincing users to share local session data or install a local tool could be a hurdle for some.
- Monetization model needs careful consideration to scale beyond individual developers to teams/enterprises.
- Competitors, though broad, could eventually build similar niche features if the market proves large enough.
Next steps
- Launch the MVP to a small group of Claude Code power users for feedback on task classification accuracy and UI.
- Validate pricing tiers, focusing on how much users would pay for the insights/savings it provides.
- Explore partnerships or integrations to access a broader user base (e.g., with developer tools, or even Anthropic if feasible for local usage data).
- Investigate methods to monitor and adapt to potential changes in Claude Code's local data formats.
- Create content demonstrating the insights gained (like the 56% example) to attract more users.
Fly Labs Method
Is the pain real, is there a gap, is it the right time, can one person build it.
- Problem clarity
- 26
- Solution gap
- 20
- Willingness to pay
- 23
- Buildability
- 12
High potential for solo builders due to a clearly articulated, high-severity problem within a specific niche, with evidence of willingness to pay and existing MVP.
Value Equation
Dream outcome and how likely it feels, against the time and effort it costs.
Strong potential for making money by solving a high-value problem for a specific, paying audience in a growing market.
One-Person Business
Curiosity pull, identity fit, and a path from free value to paid for a solo creator.
A highly focused, creator-led project solving a clear problem for a specific audience with strong monetization potential.
Viral Frameworks
Hook strength, shareability, and how cheaply it can be tested.
A strong micro-SaaS candidate with a clear target, measurable value, and a well-defined business model.
Builder Lens
Evidence the problem exists, timing, defensibility, and a model that fits on a napkin.
A highly needed, specific solution for a clear pain point, with strong evidence of immediate demand and future relevance.
Why this verdict
Five lenses, one composite. How scoring works
The angle
This weekend
- Launch the MVP to a small group of Claude Code power users for feedback on task classification accuracy and UI.
- Validate pricing tiers, focusing on how much users would pay for the insights/savings it provides.
- Explore partnerships or integrations to access a broader user base (e.g., with developer tools, or even Anthropic if feasible for local usage data).
- Investigate methods to monitor and adapt to potential changes in Claude Code's local data formats.
- Create content demonstrating the insights gained (like the 56% example) to attract more users.
Who is already there, emerging market
- HeliconeVisit
Helicone is a proxy-based observability solution that sits between your application and LLM providers, offering cost and latency visibility, caching, and automatic failover.
Pricing: Free open-source option; Hobby tier: $0; Pro plan: $20 per seat per month; Enterprise: Custom packages.
LangSmith is a unified agent engineering platform that provides observability, evaluations, and prompt engineering for any LLM application or AI agent.
Pricing: Developer plan: 1 free seat (5k base traces/month included); Plus plan: $39/month per seat (10k base traces/month included, then $0.50 per 1,000 additional base traces, $5.00 per 1,000 extended traces); Enterprise: Custom pricing.
PromptLayer is a platform for data-driven prompt engineering and LLM application management, offering prompt versioning, performance monitoring, and cost analysis.
Pricing: Free plan: 5k requests/month, 7 days logs; Pro plan: $50/user/month (unlimited log retention, up to 100,000 requests); Enterprise: Custom pricing (self-hosted option available, SOC 2 compliance).
- LiteLLMVisit
LiteLLM is an open-source proxy server that standardizes LLM calls across providers and offers cost tracking and pricing customization.
Pricing: Software license is free; infrastructure costs apply for self-hosting (estimated $200-$500/month for moderate traffic); Optional Enterprise Tier: Basic ($250/month), Premium ($30,000/year).
Vellum AI is an end-to-end LLM orchestration and observability platform designed to help teams build, deploy, and manage AI-powered apps.
Pricing: Free 'Startup' tier; Pro plan: $500/month; Enterprise: Custom pricing (annual contract, likely tens of thousands per year).
CloudZero is a cloud cost intelligence platform that provides comprehensive visibility into cloud spending, including AI costs, and attributes spend to business metrics.
Pricing: Based on annualized cloud spend, typically ranging from $10,000/year for $1M spend to $70,000+/year for enterprise-scale environments. Custom pricing for enterprise.
Anyscale is a platform for building, deploying, and scaling AI applications using the Ray framework, with a focus on distributed computing and MLOps.
Pricing: Usage-based billing; pay-as-you-go for compute instances (prices vary by CPU/GPU type and duration). Offers hosted and 'Bring Your Own Cloud' (BYOC) deployment models.
What they charge
Recent news
- 7 best free and open source LLM observability tools
PostHog, March 19 2026
- Top 9 CloudZero Alternatives & Competitors in 2026 - nOps
nOps, December 15 2025
- CloudZero Alternatives & Competitors [2026 Comparison]
nOps, April 01 2026
- Top 10 Anyscale Alternatives & Competitors in 2026 - G2
G2, 2026
- Vellum AI Pricing Guide: Is It Worth Investing In? - ZenML Blog
ZenML Blog, September 13 2025
Market signals
The market for LLM cost optimization and observability is growing rapidly, driven by increased enterprise adoption of generative AI and the need for cost control and performance monitoring. This is evidenced by numerous new tools and platforms emerging, as well as significant funding rounds for companies in the broader AI infrastructure space. The focus is shifting beyond just overall cloud spend to granular, LLM-specific metrics like token usage by task, latency, and model quality.
What frustrates people
- Unpredictable costs with usage-based billing models.
- Difficulty in connecting cloud spend to specific business outcomes or unit economics.
- Lack of deep agent tracing and built-in evaluation capabilities in some lighter-weight tools.
- Steep learning curves for complex platforms.
- Need for extensive tagging infrastructure to leverage advanced cost allocation features.
Semble – Code search for agents that uses 98% fewer tokens than grep
Hey HN! We (Stephan and Thomas) recently open-sourced Semble. We kept running into the same problem while using Claude Code on large codebases: when the agent can't find something directly, it falls back to grep, reading full files or launching subagents. This uses a lot of tokens, and often still misses the relevant code. There are existing tools for this, but they were either too slow to index on demand, needed API keys, or had poor retrieval quality.Semble is our solution for this. It combines static Model2Vec embeddings (using our latest static model: potion-code-16M) with BM25, fused via RRF and reranked with code-aware signals. Everything runs on CPU since there's no transformers involved. On our benchmark of ~1250 query/document pairs across 63 repos and 19 languages, it uses 98% fewer tokens than grep+read and reaches 99% of the retrieval quality of a 137M-parameter code-trained transformer, while being ~200x faster.Main features:- Token-efficient: 98% fewer tokens than grep+read- Fast: ~250ms to index a typical repo on our benchmark, ~1.5ms per query on CPU (very large repos may take longer)- Accurate: 0.854 NDCG@10, 99% of the best transformer setup we tested- MCP server: drop-in for Claude Code, Cursor, Codex, OpenCode- Zero config: no API keys, no GPU, no external servicesInstall in Claude Code with: claude mcp add semble -s user -- uvx --from "semble[mcp]" sembleOr check our README for other installation instructions, benchmarks, and methodology:Semble: https://github.com/MinishLab/sembleBenchmarks: https://github.com/MinishLab/semble/tree/main/benchmarksModel: https://huggingface.co/minishlab/potion-code-16MLet us know if you have any feedback or questions!
AI
Postgres extension for BM25 relevance-ranked full-text search
Last summer we faced a conundrum at my company, Tiger Data, a Postgres cloud vendor whose main business is in timeseries data. We were trying to grow our business towards emerging AI-centric workloads and wanted to provide a state-of-the-art hybrid search stack in Postgres. We'd already built pgvectorscale in house with the goal of scaling semantic search beyond pgvector's main memory limitations. We just needed a scalable ranked keyword search solution too.The problem: core Postgres doesn't provide this; the leading Postgres BM25 extension, ParadeDB, is guarded behind AGPL; developing our own extension appeared daunting. We'd need a small team of sharp engineers and 6-12 months, I figured. And we'd probably still fall short of the performance of a mature system like Parade/Tantivy.Or would we? I'd be experimenting long enough with AI-boosted development at that point to realize that with the latest tools (Claude Code + Opus) and an experienced hand (I've been working in database systems internals for 25 years now), the old time estimates pretty much go out the window.I told our CTO I thought I could solo the project in one quarter. This raised some eyebrows.It did take a little more time than that (two quarters), and we got some real help from the community (amazing!) after open-sourcing the pre-release. But I'm thrilled/exhausted today to share that pg_textsearch v1.0 is freely available via open source (Postgres license), on Tiger Data cloud, and hopefully soon, a hyperscalar near you:https://github.com/timescale/pg_textsearchIn the blog post accompanying the release, I overview the architecture and present benchmark results using MS-MARCO. To my surprise, we were not only able to meet Parade/Tantivy's query performance, but exceed it substantially, measuring a 4.7x advantage on query throughput at scale:https://www.tigerdata.com/blog/pg-textsearch-bm25-fu
AI
Consus (YC F24) – LLM gateway for companies handling sensitive data
Hey Hacker News! I'm Eric, the founder of Consus. We route all AI requests to ITAR restricted and controlled unclassified information (CUI) compliant infrastructure. Consus itself satisfies ITAR restrictions, and we recently completed our FedRAMP equivalency audit against the 410 FedRAMP High controls. FedRAMP High is the U.S. government’s strictest security baseline for cloud services handling sensitive unclassified data. A breach of this data could cause severe or catastrophic harm. Achieving FedRAMP equivalency, or authorization, is required before processing CUI information.Demo: https://youtu.be/6MYxJe6P4oYAerospace and defense is experiencing rapid growth. Unfortunately, most companies are struggling to use AI because of necessary data protections & compliance risk. These companies can move even faster if they could use the AI tools that most other industries take for granted.I spent most of my career prior to Consus in aerospace & defense. I launched national security payloads at SpaceX, and was an early employee at Castelion. I’ve watched first hand how compliance red-tape makes it difficult for companies to adopt new technologies and trust vendors that touch their sensitive data. For example, I worked with a satellite manufacturer that doesn’t allow any AI tools, even for non-sensitive work.Even as AI has become extremely useful over the past 9-12 months, I was surprised by how many aerospace & defense companies refused to touch it. They could see how it could be helpful, but they were scared about model providers training on their data, violating an ITAR rule, or risk losing new business because of a failed cyber assessment. These are all rational fears.There were a lot of ways to solve some/all of these problems, but we focused on building a compliant AI inference layer that can "plug" into any AI tool or workflow. Just as OpenRouter integrates with most tools, Consus can connect to tools like ChatGPT or Clau
AI
Questions about this idea?
FlyBot reads the scoring and gives you a second opinion on “CodeBurn – Analyze Claude Code token usage by task”.