tracellm 0.2.0__tar.gz → 0.2.2__tar.gz

{tracellm-0.2.0 → tracellm-0.2.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: tracellm
-Version: 0.2.0
+Version: 0.2.2
 Summary: Local-first observability for LLM applications
 Requires-Python: >=3.9
 Description-Content-Type: text/markdown
@@ -13,45 +13,78 @@ every API interaction logged locally, queryable from your terminal.
 No backend. No signup. Nothing leaves your machine.
 
 ## Install
-
+```bash
 pip install tracellm
+```
 
 ## Usage
-
+```python
 from tracellm import trace
 import groq
 
 client = groq.Groq(api_key="your-key")
 
 @trace
-def llm_call(model="llama-3.1-8b-instant", messages=[{"role": "user", "content": "hello"}]):
+def ask(model, messages):
     return client.chat.completions.create(model=model, messages=messages)
 
-llm_call()
-
-That's it. Every call is traced automatically.
+ask(
+    model="llama-3.1-8b-instant",
+    messages=[{"role": "user", "content": "Explain black holes in one line"}]
+)
+```
 
-## Query traces from terminal
+Every call is traced automatically. No try/except. No setup.
 
+## Query traces
+```bash
 python -m tracellm.cli --Status failed
 python -m tracellm.cli --Latency 2.0
 python -m tracellm.cli --Model llama-3.1-8b-instant
 python -m tracellm.cli --Status failed --Latency 1.5
+python -m tracellm.cli --Time "2026-04-03"
+```
 
-## What gets captured
+## Cost tracking
+```bash
+# cost per trace
+python -m tracellm.cli --Cost
+
+# full summary by model
+python -m tracellm.cli --Cost Summary
+```
+
+Output:
+=== Cost Summary ===
+llama-3.1-8b-instant
+Calls  : 8
+Tokens : 405
+Cost   : $0.000020
 
+Total calls made : 8
+Total tokens used: 405
+Total cost       : $0.000020
+
+## What gets captured
 - Model, prompt, response
 - Tokens used, latency, finish reason
 - Error type and message on failures
 - Timestamp for every call
 
-## Limitations
+## Pricing
+Default pricing is bundled. To override, create `~/.tracellm/pricing.json`:
+```json
+{
+  "my-custom-model": 0.05
+}
+```
+Values are per million tokens.
 
-Storage is append-only JSON lines. Latency query supports >=
-for latency, exact match for everything else. Early days.
+## Limitations
+Storage is append-only JSON lines. Latency filter supports `>=`,
+exact match for everything else. Early days.
 
 ## Roadmap
-
 - Binary storage for faster querying at scale
-- Cost calculation per model
+- Async tracing support
 - Terminal dashboard

tracellm-0.2.2/README.md

@@ -0,0 +1,83 @@
+# tracellm
+
+Lightweight tracing for LLM applications. One decorator —
+every API interaction logged locally, queryable from your terminal.
+
+No backend. No signup. Nothing leaves your machine.
+
+## Install
+```bash
+pip install tracellm
+```
+
+## Usage
+```python
+from tracellm import trace
+import groq
+
+client = groq.Groq(api_key="your-key")
+
+@trace
+def ask(model, messages):
+    return client.chat.completions.create(model=model, messages=messages)
+
+ask(
+    model="llama-3.1-8b-instant",
+    messages=[{"role": "user", "content": "Explain black holes in one line"}]
+)
+```
+
+Every call is traced automatically. No try/except. No setup.
+
+## Query traces
+```bash
+python -m tracellm.cli --Status failed
+python -m tracellm.cli --Latency 2.0
+python -m tracellm.cli --Model llama-3.1-8b-instant
+python -m tracellm.cli --Status failed --Latency 1.5
+python -m tracellm.cli --Time "2026-04-03"
+```
+
+## Cost tracking
+```bash
+# cost per trace
+python -m tracellm.cli --Cost
+
+# full summary by model
+python -m tracellm.cli --Cost Summary
+```
+
+Output:
+=== Cost Summary ===
+llama-3.1-8b-instant
+Calls  : 8
+Tokens : 405
+Cost   : $0.000020
+
+Total calls made : 8
+Total tokens used: 405
+Total cost       : $0.000020
+
+## What gets captured
+- Model, prompt, response
+- Tokens used, latency, finish reason
+- Error type and message on failures
+- Timestamp for every call
+
+## Pricing
+Default pricing is bundled. To override, create `~/.tracellm/pricing.json`:
+```json
+{
+  "my-custom-model": 0.05
+}
+```
+Values are per million tokens.
+
+## Limitations
+Storage is append-only JSON lines. Latency filter supports `>=`,
+exact match for everything else. Early days.
+
+## Roadmap
+- Binary storage for faster querying at scale
+- Async tracing support
+- Terminal dashboard

{tracellm-0.2.0 → tracellm-0.2.2}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "tracellm"
-version = "0.2.0"
+version = "0.2.2"
 description = "Local-first observability for LLM applications"
 readme = "README.md"
 requires-python = ">=3.9"

tracellm-0.2.2/tracellm/__init__.py

@@ -0,0 +1,3 @@
+from .tracer import Tracer
+from .decorator import trace
+from .pricing import load_pricing

tracellm-0.2.2/tracellm/cli.py

@@ -0,0 +1,108 @@
+import argparse
+import json
+import os
+from datetime import datetime
+from tracellm import load_pricing
+
+TRACE_FILE = os.path.join(os.path.dirname(__file__), '..', 'trace.txt')
+
+Running_total = {"Model" : None , "Calls" : 0 , "Tokens" : 0 , "Cost" : 0}
+
+parser = argparse.ArgumentParser()
+parser.add_argument("-S" , "--Status" , help = "Based on what status do filter" , choices = ["success" , "failed"])
+parser.add_argument("-L" , "--Latency" , help = "Based on what latency do filter")
+parser.add_argument("-M" , "--Model" , help = "Based on what model do filter")
+parser.add_argument("-E" , "--Error" , help = "Based on what type of Error do filter")
+parser.add_argument("-T" , "--Time" , help = "Based on what time do filter")
+parser.add_argument("-C" , "--Cost" , help = "Calculates the token cost , use word 'Summary' for cost summary")
+args = parser.parse_args()
+
+def parse_time_input(s):
+    for fmt in ("%Y-%m-%d %H:%M:%S", "%Y-%m-%d"):
+        try:
+            return datetime.strptime(s, fmt)
+        except ValueError:
+            continue
+    raise ValueError(f"Time format not recognised: {s}")
+conditions = []
+if args.Status:
+    conditions.append(lambda t : t.get('Status') == args.Status)
+if args.Latency:
+    conditions.append(lambda t : t.get('Latency') >= float(args.Latency))
+if args.Model:
+    conditions.append(lambda t : t.get('Model') == args.Model)
+if args.Error:
+    conditions.append(lambda t : t.get('Error Type' , None) == args.Error)
+if args.Time:
+    query_time = parse_time_input(args.Time)
+    conditions.append(
+        lambda t, qt=query_time: datetime.strptime(
+            t.get('Timestamp', ''), "%Y-%m-%d %H:%M:%S"
+        ) >= qt
+    )
+if conditions != []:
+    with open(TRACE_FILE , 'r') as trace_file:
+        for line in trace_file:
+            trace = json.loads(line)
+            if all(condition(trace) for condition in conditions):
+                print("\n--- Trace ---")
+                for key, value in trace.items():
+                    print(f"  {key}: {value}")
+                print("-------------")
+            
+if args.Cost:
+    pricing = load_pricing()
+    total_calls = 0
+    total_tokens = 0
+    total_cost = 0
+    model_breakdown = {}
+    if args.Cost != 'Summary':
+        C_or_S = 'Cost'
+    else:
+        C_or_S = 'Summary'
+    with open(TRACE_FILE , 'r') as f:
+        for line in f:
+            trace = json.loads(line)
+            if 'Tokens used' not in trace:
+                continue
+            T_used = trace['Tokens used']
+            if trace['Model'] not in pricing:
+                print(f"Model :{trace['Model']} is not stored.\nPlease config the pricing file and add the model ")
+                continue
+            cost = T_used * pricing[trace['Model']] / 1000000
+            if C_or_S == 'Cost':
+                print(f"Model : {trace['Model']}")
+                print(f"Tokens Used : {T_used}")
+                print(f"Cost : {cost}")
+                print("---------------------\n")
+            else:
+                total_cost += cost
+                total_calls += 1
+                total_tokens += trace['Tokens used']
+                if trace['Model'] in model_breakdown:
+                    model_breakdown[trace['Model']]['Total Calls'] += 1
+                    model_breakdown[trace['Model']]['Total Tokens'] += trace['Tokens used']
+                    model_breakdown[trace['Model']]['Total Cost'] += cost
+                else:
+                    model_breakdown[trace['Model']] = {}
+                    model_breakdown[trace['Model']] = {
+                        'Total Calls' : 1,
+                        'Total Tokens' : trace['Tokens used'],
+                        'Total Cost' : cost
+                    }
+        if C_or_S == 'Summary':
+            print("\n=== Cost Summary ===")
+            for model, data in model_breakdown.items():
+                print(f"\n  {model}")
+                print(f"    Calls  : {data['Total Calls']}")
+                print(f"    Tokens : {data['Total Tokens']}")
+                print(f"    Cost   : ${data['Total Cost']:.6f}")
+                print("====================")
+            print(f"\n     Total calls made : {total_calls}")
+            print(f"     Total Tokens used : {total_tokens}")
+            print(f"     Total cost: ${total_cost:.6f}")
+            print("====================")
+                
+
+
+

tracellm-0.2.2/tracellm/pricing.py

@@ -0,0 +1,33 @@
+import os
+import json
+Pricing = {
+    "gpt-oss-20B (low)": 0.0675,
+    "gpt-oss-20B (high)": 1.25,
+    "Gemini 3.1 Pro Preview": 1.15,
+    "GPT-5.4 (xhigh)": 12.50,
+    "GPT-5.3 Codex (xhigh)": 11.50,
+    "Claude Opus 4.6 (max)": 11.25,
+    "Mercury 2": 0.04,
+    "Qwen3.5 0.8B": 0.02,
+    "Granite 4.0 H Small": 0.03,
+    "Granite 3.3 8B": 0.04,
+    "DeepSeek R1 Distill Qwen 32B": 0.09,
+    "NVIDIA Nemotron 3 Nano": 0.03,
+    "Gemma 3n E4B": 0.03,
+    "Qwen3.5 2B": 0.04,
+    "Llama 4 Scout": 0.05,
+    "Grok 4.20 Beta 0309": 0.45,
+    "Gemini 2.0 Pro Experimental": 1.10,
+    "llama-3.1-8b-instant": 0.05
+}
+
+user_home = os.path.expanduser('~')
+config_dirc = os.path.join(user_home , '.tracellm')
+personal_dict = os.path.join(config_dirc , 'pricing.json')
+def load_pricing():
+    if os.path.exists(personal_dict):
+        with open(personal_dict , 'r') as f:
+            return json.load(f)
+    else:
+        return Pricing
+    

{tracellm-0.2.0 → tracellm-0.2.2}/tracellm/tracer.py

@@ -2,6 +2,7 @@ import json
 import time
 from datetime import datetime
 import os
+
 TRACE_FILE = os.path.join(os.path.dirname(__file__), '..', 'trace.txt')
 
 class Tracer:

{tracellm-0.2.0 → tracellm-0.2.2}/tracellm.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: tracellm
-Version: 0.2.0
+Version: 0.2.2
 Summary: Local-first observability for LLM applications
 Requires-Python: >=3.9
 Description-Content-Type: text/markdown
@@ -13,45 +13,78 @@ every API interaction logged locally, queryable from your terminal.
 No backend. No signup. Nothing leaves your machine.
 
 ## Install
-
+```bash
 pip install tracellm
+```
 
 ## Usage
-
+```python
 from tracellm import trace
 import groq
 
 client = groq.Groq(api_key="your-key")
 
 @trace
-def llm_call(model="llama-3.1-8b-instant", messages=[{"role": "user", "content": "hello"}]):
+def ask(model, messages):
     return client.chat.completions.create(model=model, messages=messages)
 
-llm_call()
-
-That's it. Every call is traced automatically.
+ask(
+    model="llama-3.1-8b-instant",
+    messages=[{"role": "user", "content": "Explain black holes in one line"}]
+)
+```
 
-## Query traces from terminal
+Every call is traced automatically. No try/except. No setup.
 
+## Query traces
+```bash
 python -m tracellm.cli --Status failed
 python -m tracellm.cli --Latency 2.0
 python -m tracellm.cli --Model llama-3.1-8b-instant
 python -m tracellm.cli --Status failed --Latency 1.5
+python -m tracellm.cli --Time "2026-04-03"
+```
 
-## What gets captured
+## Cost tracking
+```bash
+# cost per trace
+python -m tracellm.cli --Cost
+
+# full summary by model
+python -m tracellm.cli --Cost Summary
+```
+
+Output:
+=== Cost Summary ===
+llama-3.1-8b-instant
+Calls  : 8
+Tokens : 405
+Cost   : $0.000020
 
+Total calls made : 8
+Total tokens used: 405
+Total cost       : $0.000020
+
+## What gets captured
 - Model, prompt, response
 - Tokens used, latency, finish reason
 - Error type and message on failures
 - Timestamp for every call
 
-## Limitations
+## Pricing
+Default pricing is bundled. To override, create `~/.tracellm/pricing.json`:
+```json
+{
+  "my-custom-model": 0.05
+}
+```
+Values are per million tokens.
 
-Storage is append-only JSON lines. Latency query supports >=
-for latency, exact match for everything else. Early days.
+## Limitations
+Storage is append-only JSON lines. Latency filter supports `>=`,
+exact match for everything else. Early days.
 
 ## Roadmap
-
 - Binary storage for faster querying at scale
-- Cost calculation per model
+- Async tracing support
 - Terminal dashboard

{tracellm-0.2.0 → tracellm-0.2.2}/tracellm.egg-info/SOURCES.txt

@@ -3,6 +3,7 @@ pyproject.toml
 tracellm/__init__.py
 tracellm/cli.py
 tracellm/decorator.py
+tracellm/pricing.py
 tracellm/tracer.py
 tracellm.egg-info/PKG-INFO
 tracellm.egg-info/SOURCES.txt

tracellm-0.2.0/README.md

@@ -1,50 +0,0 @@
-# tracellm
-
-Lightweight tracing for LLM applications. One decorator —
-every API interaction logged locally, queryable from your terminal.
-
-No backend. No signup. Nothing leaves your machine.
-
-## Install
-
-pip install tracellm
-
-## Usage
-
-from tracellm import trace
-import groq
-
-client = groq.Groq(api_key="your-key")
-
-@trace
-def llm_call(model="llama-3.1-8b-instant", messages=[{"role": "user", "content": "hello"}]):
-    return client.chat.completions.create(model=model, messages=messages)
-
-llm_call()
-
-That's it. Every call is traced automatically.
-
-## Query traces from terminal
-
-python -m tracellm.cli --Status failed
-python -m tracellm.cli --Latency 2.0
-python -m tracellm.cli --Model llama-3.1-8b-instant
-python -m tracellm.cli --Status failed --Latency 1.5
-
-## What gets captured
-
-- Model, prompt, response
-- Tokens used, latency, finish reason
-- Error type and message on failures
-- Timestamp for every call
-
-## Limitations
-
-Storage is append-only JSON lines. Latency query supports >=
-for latency, exact match for everything else. Early days.
-
-## Roadmap
-
-- Binary storage for faster querying at scale
-- Cost calculation per model
-- Terminal dashboard

tracellm-0.2.0/tracellm/__init__.py

@@ -1,2 +0,0 @@
-from .tracer import Tracer
-from .decorator import trace

tracellm-0.2.0/tracellm/cli.py

@@ -1,33 +0,0 @@
-import argparse
-import json
-import os
-TRACE_FILE = os.path.join(os.path.dirname(__file__), '..', 'trace.txt')
-
-parser = argparse.ArgumentParser()
-parser.add_argument("-S" , "--Status" , help = "Based on what status do filter" , choices = ["success" , "failed"])
-parser.add_argument("-L" , "--Latency" , help = "Based on what latency do filter")
-parser.add_argument("-M" , "--Model" , help = "Based on what model do filter")
-parser.add_argument("-E" , "--Error" , help = "Based on what type of Error do filter")
-parser.add_argument("-T" , "--Time" , help = "Based on what time do filter")
-args = parser.parse_args()
-
-
-conditions = []
-if args.Status:
-    conditions.append(lambda t : t.get('Status') == args.Status)
-if args.Latency:
-    conditions.append(lambda t : t.get('Latency') >= float(args.Latency))
-if args.Model:
-    conditions.append(lambda t : t.get('Model') == args.Model)
-if args.Error:
-    conditions.append(lambda t : t.get('Error Type' , None) == args.Error)
-if args.Time:
-    conditions.append(lambda t : t.get('Start Time') >= float(args.Time))
-with open(TRACE_FILE , 'r') as trace_file:
-    for line in trace_file:
-        trace = json.loads(line)
-        if all(condition(trace) for condition in conditions):
-            print("\n--- Trace ---")
-            for key, value in trace.items():
-                print(f"  {key}: {value}")
-            print("-------------")