Documentation

API Reference

Everything you need to run neuromorphic simulations from any language.

Base URL: https://api.catalyst-neuromorphic.com

Python SDK (recommended)

pip install catalyst-cloud

import catalyst_cloud as cc

client = cc.Client("cn_live_...")
net = client.create_network(
    populations=[{"label": "input", "size": 100, "params": {"threshold": 1000}}],
    connections=[],
)
job = client.simulate(net["network_id"], timesteps=500,
    stimuli=[{"population": "input", "current": 5000}])
print(f"Spikes: {job['result']['total_spikes']}")

PyPI · GitHub · Interactive demo

Quick start (REST API)

1. Get an API key

Free. No credit card required.

curl -X POST https://api.catalyst-neuromorphic.com/v1/signup \
  -H "Content-Type: application/json" \
  -d '{"email": "[email protected]", "tier": "free"}'

# Response:
# {
#   "api_key": "cn_live_...",
#   "tier": "free",
#   "limits": {"max_neurons": 1024, "max_timesteps": 1000, ...}
# }

import requests

r = requests.post("https://api.catalyst-neuromorphic.com/v1/signup",
    json={"email": "[email protected]", "tier": "free"})

data = r.json()
API_KEY = data["api_key"]  # Save this — shown only once
print(f"Key: {data['key_prefix']}...")
print(f"Limits: {data['limits']}")

const res = await fetch("https://api.catalyst-neuromorphic.com/v1/signup", {
  method: "POST",
  headers: { "Content-Type": "application/json" },
  body: JSON.stringify({ email: "[email protected]", tier: "free" })
});

const { api_key, limits } = await res.json();
// Save api_key — it's shown only once

2. Define a network

Populations of spiking neurons with connections between them.

curl -X POST https://api.catalyst-neuromorphic.com/v1/networks \
  -H "Content-Type: application/json" \
  -H "X-API-Key: cn_live_..." \
  -d '{
    "populations": [
      {"label": "excitatory", "size": 100, "params": {"threshold": 1000, "leak": 10}},
      {"label": "inhibitory", "size": 25, "params": {"threshold": 800, "leak": 5}}
    ],
    "connections": [
      {"source": "excitatory", "target": "inhibitory", "topology": "random_sparse",
       "p": 0.2, "weight": 600},
      {"source": "inhibitory", "target": "excitatory", "weight": -400}
    ]
  }'

# Response:
# {"network_id": "net_...", "total_neurons": 125, "populations": 2, "connections": 2}

headers = {"X-API-Key": API_KEY, "Content-Type": "application/json"}

network = {
    "populations": [
        {"label": "excitatory", "size": 100, "params": {"threshold": 1000, "leak": 10}},
        {"label": "inhibitory", "size": 25, "params": {"threshold": 800, "leak": 5}},
    ],
    "connections": [
        {"source": "excitatory", "target": "inhibitory",
         "topology": "random_sparse", "p": 0.2, "weight": 600},
        {"source": "inhibitory", "target": "excitatory", "weight": -400},
    ],
}

r = requests.post("https://api.catalyst-neuromorphic.com/v1/networks",
    json=network, headers=headers)

net_id = r.json()["network_id"]

const headers = {
  "X-API-Key": api_key,
  "Content-Type": "application/json"
};

const res = await fetch("https://api.catalyst-neuromorphic.com/v1/networks", {
  method: "POST", headers,
  body: JSON.stringify({
    populations: [
      { label: "excitatory", size: 100, params: { threshold: 1000, leak: 10 } },
      { label: "inhibitory", size: 25, params: { threshold: 800, leak: 5 } }
    ],
    connections: [
      { source: "excitatory", target: "inhibitory",
        topology: "random_sparse", p: 0.2, weight: 600 },
      { source: "inhibitory", target: "excitatory", weight: -400 }
    ]
  })
});

const { network_id } = await res.json();

3. Run a simulation

Submit a job with stimuli. It runs asynchronously and you poll for results.

curl -X POST https://api.catalyst-neuromorphic.com/v1/jobs \
  -H "Content-Type: application/json" \
  -H "X-API-Key: cn_live_..." \
  -d '{
    "network_id": "net_...",
    "timesteps": 1000,
    "stimuli": [
      {"population": "excitatory", "neurons": [0,1,2,3,4], "current": 8000}
    ]
  }'

# Response: {"job_id": "job_...", "status": "queued"}

# Poll for results:
curl https://api.catalyst-neuromorphic.com/v1/jobs/job_... \
  -H "X-API-Key: cn_live_..."

# When status is "completed":
# {
#   "status": "completed",
#   "result": {
#     "total_spikes": 847,
#     "firing_rates": {"excitatory": 0.0052, "inhibitory": 0.0128},
#     "spike_count_timeseries": [5, 12, 3, ...]
#   },
#   "compute_seconds": 0.34
# }

import time

# Submit job
r = requests.post("https://api.catalyst-neuromorphic.com/v1/jobs",
    json={
        "network_id": net_id,
        "timesteps": 1000,
        "stimuli": [{"population": "excitatory", "neurons": [0,1,2,3,4], "current": 8000}],
    },
    headers=headers)

job_id = r.json()["job_id"]

# Poll for completion
while True:
    r = requests.get(f"https://api.catalyst-neuromorphic.com/v1/jobs/{job_id}",
        headers=headers)
    job = r.json()
    if job["status"] in ("completed", "failed"):
        break
    time.sleep(1)

print(f"Spikes: {job['result']['total_spikes']}")
print(f"Rates: {job['result']['firing_rates']}")
print(f"Compute: {job['compute_seconds']}s")

// Submit job
let res = await fetch("https://api.catalyst-neuromorphic.com/v1/jobs", {
  method: "POST", headers,
  body: JSON.stringify({
    network_id,
    timesteps: 1000,
    stimuli: [{ population: "excitatory", neurons: [0,1,2,3,4], current: 8000 }]
  })
});

const { job_id } = await res.json();

// Poll for completion
let job;
do {
  await new Promise(r => setTimeout(r, 1000));
  res = await fetch(`https://api.catalyst-neuromorphic.com/v1/jobs/${job_id}`, { headers });
  job = await res.json();
} while (job.status !== "completed" && job.status !== "failed");

console.log("Spikes:", job.result.total_spikes);
console.log("Rates:", job.result.firing_rates);

4. Get spike trains

Full per-neuron spike times, indexed by population label and local neuron index.

curl https://api.catalyst-neuromorphic.com/v1/jobs/job_.../spikes \
  -H "X-API-Key: cn_live_..."

# Response:
# {
#   "spike_trains": {
#     "excitatory": {"0": [12, 45, 78], "1": [13, 47], "4": [11, 44, 77, 110]},
#     "inhibitory": {"0": [14, 48], "3": [15, 49, 82]}
#   }
# }

r = requests.get(f"https://api.catalyst-neuromorphic.com/v1/jobs/{job_id}/spikes",
    headers=headers)

spikes = r.json()["spike_trains"]

# Plot raster
import matplotlib.pyplot as plt

fig, ax = plt.subplots(figsize=(10, 4))
y = 0
for pop, neurons in spikes.items():
    for nid, times in sorted(neurons.items(), key=lambda x: int(x[0])):
        ax.scatter(times, [y] * len(times), s=1, c='white')
        y += 1
ax.set_xlabel("Timestep")
ax.set_ylabel("Neuron")
plt.savefig("raster.png", dpi=150)
print("Saved raster.png")

Endpoint reference

POST /v1/signup Public

Create a new account and get an API key.

Request body:

email	string, required
tier	"free" (default), "researcher", "startup", "enterprise"

POST /v1/networks Requires X-API-Key

Store a network definition for later job submission.

Request body:

populations	array of label, size, params
connections	array of source, target, topology, weight, p, delay, compartment

Neuron params (all optional):

threshold	int (default 1000) — spike threshold
leak	int (default 3) — membrane leak rate
resting	int (default 0) — resting potential
refrac	int (default 3) — refractory period
dend_threshold	int (default 500) — dendritic compartment threshold

Connection options:

topology	all_to_all, one_to_one, random_sparse, fixed_fan_in, fixed_fan_out
weight	int [-32768, 32767] (default 200) — negative for inhibitory
p	float [0, 1] — connection probability (random_sparse)
delay	int [0, 63] — synaptic delay in timesteps
compartment	int [0, 3] — target dendritic compartment

POST /v1/jobs Requires X-API-Key

Submit a simulation job. Returns immediately with job_id.

network_id	string, required — from POST /v1/networks
timesteps	int, required — number of simulation steps
stimuli	array of population, neurons (optional), current
learning	enabled, graded, dendritic, noise, three_factor
rewards	array of timestep, value — for 3-factor learning

GET /v1/jobs/job_id Requires X-API-Key

Get job status and summary results. Poll until status is "completed" or "failed".

Response fields (when completed):

result.total_spikes	int — total spikes across all neurons
result.firing_rates	per-population mean spikes per neuron per timestep
result.spike_count_timeseries	array — total spikes at each timestep
compute_seconds	float — actual simulation compute time

GET /v1/jobs/job_id/spikes Requires X-API-Key

Full spike trains grouped by population with local neuron indices.

Response:

{"spike_trains": {"excitatory": {"0": [12, 45], "1": [13]}, "inhibitory": {"0": [14, 48]}}}

GET /v1/usage Requires X-API-Key

Your current usage stats and estimated cost.

jobs_today	int — jobs submitted today
jobs_limit_today	int — daily limit for your tier
compute_seconds_this_month	float — total compute this billing cycle
estimated_cost_gbp	float — estimated cost at your tier rate

Error codes

400	Invalid request — check your JSON format
401	Missing or invalid API key
403	Tier limit exceeded (neurons, timesteps)
404	Network or job not found
409	Email already registered
429	Daily job limit reached — upgrade your tier

Interactive docs also available at

OpenAPI Explorer →