One key for
every coding agent.
Claude Code and Codex CLI both work against ObixHub unchanged. Point the base URL here, drop in your obx_ key, and put Claude and GPT on one flat monthly plan — with a generous usage allowance, your whole team’s usage in one dashboard, and no stored prompts.
Three steps. Sixty seconds.
Set one env var to https://obixhub.com. Each SDK appends its own path — no /v1 to remember, no plugin to install.
One named key, scoped to your team’s plan and an optional per-member usage limit. Use it on every endpoint.
That’s the whole setup. Your agent runs exactly as before — same commands, same workflow — now backed by every Claude and GPT model.
Your CLI. One base URL.
Each works as-is — no plugin, no fork. Point at the one base URL below, drop in your obx_ key, done.
export ANTHROPIC_BASE_URL="https://obixhub.com"
export ANTHROPIC_API_KEY="obx_ER0B******dzKp"
$ claude# ~/.codex/config.toml
model_provider = "obixhub"
[model_providers.obixhub]
base_url = "https://obixhub.com/v1"
wire_api = "responses"
env_key = "OPENAI_API_KEY"
export OPENAI_API_KEY="obx_ER0B******dzKp"
$ codexOne key where you’d otherwise juggle two.
One flat monthly price. No surprises.
Pick a plan and run Claude and GPT against it all month — no per-token metering, no overage shock. Four tiers, from $25 to $239 a month, each with a generous weekly usage allowance.
- basic · $25/mo≈ $80/wk usage25 spots left
- plus · $79/mo≈ $200/wk usage10 spots left
- pro · $159/mo≈ $420/wk usage4 spots left
- max · $239/mo≈ $650/wk usage3 spots left
- paymentusd card · stripe
- A flat monthly bill — the same number every cycle, from $25 to $239.
- Pro packs about $1680/mo of usage at official Claude or GPT rates.
- Claude and GPT on one key — switch models by changing one string.
- Upgrade, downgrade, or cancel anytime from your dashboard.
Plenty of headroom. One dashboard.
One plan covers the team, sized for real daily work. Every member gets their own key — and you see exactly where the allowance goes.
Every tier ships a generous weekly usage allowance — enough to run Claude and GPT through a full week of daily coding, no rationing, no overage shock.
One subscription covers the whole team. Add members, give each their own named key and model allowlist, and manage the plan from a single dashboard.
Track usage by member, model, and time — requests, tokens, and how much of the weekly allowance each member has used. Metadata only; never your prompts.
Your prompts and completions never land in our storage.
We record only metadata — token counts, latency, model, and status. Never message bodies, never tool inputs. Your conversations pass through and are gone.
And we route every request through the providers’ zero-data-retention paths, so your prompts aren’t kept upstream either — not by us, not by the model providers.
Questions, answered.
Which tools work?
Claude Code and Codex CLI work unchanged — and any other client that speaks the Anthropic Messages or OpenAI Chat/Responses wire format and lets you set its base URL connects the same way.
Do you store my prompts?
No. We record only metadata — token counts, latency, model, and status. Message bodies and tool inputs pass through and are never written to our storage, and we route through the providers’ zero-data-retention paths so they don’t retain them either.
How am I billed?
A flat monthly subscription, in USD by card — with an invoice every cycle, so it’s easy to expense. Pick a plan and use Claude and GPT against it all month: no per-token metering, no surprise bill, and a generous weekly usage allowance.
Can I switch models without changing code?
Yes. The same key reaches every model in the catalog — move between Claude and GPT by changing the model string.
Can my team share one plan?
Yes. A team holds one subscription; every member gets their own named key and an optional usage limit, all drawing from the shared plan — managed from one dashboard.
How much usage does a plan include?
Each tier comes with a generous weekly usage allowance, sized so a team can run Claude and GPT through a normal week of daily coding. Higher tiers carry a larger allowance, and you can see how much is left at any time in the dashboard.
Can I see how the plan is being used?
Yes. Your dashboard breaks usage down by member, model, and time — requests, tokens, and how much of the weekly allowance you’ve used — so you always know where the plan is going. It’s usage metadata only; we never store your prompts.
Can I change or cancel my plan?
Yes. Upgrade, downgrade, or cancel anytime from your dashboard — changes take effect on your next billing cycle.
Pick a plan. Get a key. Ship today.
≈ 60 seconds from signup → first stream