Ai

Helios Pro workspace

AI Dashboard

Model usage, token economics and inference health across the Helios Pro AI gateway.

Total Calls

6.45M

24h window

+18.2%vs last month

Tokens Processed

3.24B

input + output

+22.4%vs last month

Avg. Latency

284ms

p50 across models

-38msvs last month

Inference Cost

$32,900

$0.0051 / 1k tokens

+4.1%vs last month

Model Usage

Daily calls, last 14 days

+18.2%

Tokens by Category

Where tokens are spent

Latency Trend

p50 / p95 / p99 (ms)

-38ms p50

Live Inference

Gateway event stream

Live
  1. AS

    Aarav Sharma ran inference on Helios-2 Turbo (4.2k tokens)

    12 sec ago
  2. PP

    Priya Patel fine-tuned Llama 3.1 70B

    2 min ago
  3. DC

    Daniel Chen deployed Mistral Large v3

    8 min ago
  4. SM

    Sofia Mendes hit rate limit on Claude 3.5 Sonnet

    14 min ago
  5. MB

    Marcus Bell optimized prompt for GPT-4o

    26 min ago

Model Performance

Calls, tokens, latency, cost & health

6 models
ModelCallsTokensLatencyCostHealth

Helios-2 Turbo

Helios Labs

2.4M1.2B184ms
$8,420
healthy

GPT-4o

OpenAI

1.8M920M412ms
$12,400
healthy

Claude 3.5 Sonnet

Anthropic

940k480M396ms
$6,820
degraded

Llama 3.1 70B

Self-hosted

620k320M248ms
$1,240
healthy

Mistral Large

Mistral AI

410k180M322ms
$2,180
down

Gemini 1.5 Pro

Google

280k140M288ms
$1,840
healthy